Dejan Bulić, The Fortifications of The Late Antiquity and The Early Byzantine Period On The Later Territory of The South-Slavic Principalities, and Their Re-Occupation

THE WORLD OF THE SLAVS
Studies on the East, West and South Slavs:

Civitas, Oppidas, Villas and Archeological Evidence
(7th to 11th Centuries AD)
Reviewers:
Academician Jovanka Kali
Prof. Vlada Stankovi
Assist. Prof. Dejan Radievi
Milan Radujko, Ph.D.
Lovorka orali, Ph.D.
This book has been published with the financial support of

THE MINISTRY OF THE EDUCATION, SCIENCE AND TEHNOLOGICAL
DEVELOPMENT OF THE REPUBLIC OF SERBIA
(project No III 47025)
THE INSTITUTE OF HISTORY

Monographs
Vol. 64
Tibor ivkovi Dejan Crnevi, Dejan Buli

Vladeta Petrovi, Irena Cvijanovi, Bojana Radovanovi
THE WORLD OF THE SLAVS

Studies on the East, West and South Slavs:
Civitas, Oppidas, Villas and Archeological Evidence
(7th to 11th Centuries AD)
Editor in chief
Sran Rudi, Ph.D.
Director of The Institute of History
Belgrade
2013
TA B L E O F C O N T E N TS
9
13
Preface
List of Abbreviations
15
The Urban Landcape of Early Medieval Slavic Principalities in

the Territories of the Former Praefectura Illyricum and in the
Province of Dalmatia (ca. 610 950)
TIBOR IVKOVI
37
The Architecture of Cathedral Churches on the Eastern Adriatic

Coast at the Time of the First Principalities of South Slavs (9th 11th
Centuries)
DEJAN CRNEVI
137
The Fortifications of the Late Antiquity and the Early Byzantine

Period on the Later Territory of the South-Slavic Principalities,
and Their Re-occupation
DEJAN BULI
235
Terrestrial Communications in the Late Antiquity and the Early

Middle Ages in the Western Part of the Balkan Peninsula
VLADETA PETROVI
289
The Typology of Early Medieval Settlements in Bohemia, Poland

and Russia
IRENA CVIJANOVI
345
The Typology of Slavic Settlements in Central Europe in the

Middle Ages According to Latin Sources (8th 12th Centuries)
BOJANA RADOVANOVI
369
413
421
429
List of References
List of Illustrations
General Index
The Fortifications of the Late Antiquity and the Early

Byzantine Period on the Later Territory of the SouthSlavic Principalities, and Their Re-occupation
DEJAN BULI
As the title reveals, this text will cover the Early Byzantine period
(early 5th early 7th century) in the areas we have surveyed ourselves, i.e.
Serbia, Croatia and Montenegro. However, as some authors use the two
designations the Late Antiquity and the Early Byzantine period synonymously, the time frame for the territories of Bosnia and Herzegovina, and
Macedonia is set from the 330s to 610s. It was already pointed out that a
precise chronological estimate cannot be determined without excavation
works and analysis of ceramics and small findings.184 The sites indexed
with poor, often just unspecific data, and described in acquired, conservative
interpretations, offer insecure datings, making fine-tuned chronological
estimations impossible, most of the time. For all these reasons, a revision
of the already-existing lists of sites for the territories of Macedonia and
Bosnia and Herzegovina could not be done, as the material was impossible
to gain insight into.
Considering territory, the work will cover the area of the former
Yugoslavia, without Slovenia and Istria, or more precisely, the area delimited by the river Raa on the north, i. e. Croatia, Bosnia and Herzegovina,
Montenegro, Serbia, and Macedonia. In other words, the territories that
184
M. Garaanin, Odbrambeni sistemi u praistoriji i antici na tlu Jugoslavije,

Materijali 22, Novi Sad 1986, 10; I. remonik, Rimska utvrenja u BiH s
osvrtom na utvrenja kasne antike, Arheoloki Vestnik 41 (1990) 355 (=
remonik, Rimska utvrenja).
138
Dejan Buli
first formed a part of South Slavic principalities, and then states, during the
Early Middle Ages. During the research undertaken until now in this area,
a large number of fortifications were noticed, with a cultural layer from
the Late Antiquity and Early Byzantine period. Information on these was,
in a large measure, obtained through terrain reconnaissance. Sondages
were undertaken on dozens of sites, while systematic archaeological excavations were seldom conducted. The territory covered in this work
encompasses geographical entities defined according to the present-day
administrative borders of the states, which is why we did not take into
account the provincial demarcations from the Late Antiquity/Early
Byzantine period, as these present-day territories were part of two, three
or several provinces throughout the Late Antiquity.185
After the creation of lists and maps of the Late Antiquity/Early
Byzantine localities, the final objective of this work is registering the Early
Medieval and Medieval strata in the mentioned fortifications, and on
detecting potential continuity and discontinuity that marked the medieval
and Early Byzantine period. It is difficult to report some of the relevant
data about the construction or particularities of specific fortifications, their
functions, interconnections, and the roles they played in the defence system
of the Empire in the Late Antiquity or the Early Byzantine era. The aim of this
research is to reflect on the historical context, and not on the movable archaeological material, which is a task beyond the scope of this kind of study.
Some zones of present-day countries remain insufficiently covered,
a consequence of local museums policies and interest, because of which
some areas have not even been reconnoitred, which caused uneven level
of exploration among the regions. For example, the regions of continental
Croatia are the least examined territory.
All that was mentioned above pertains to medieval sites, too, and to
a far greater degree, as they were neglected. They were not the subjects of
initiated projects, but have always remained out of the focus of researchers
to such a degree that these days clear distinction between the Late
Antiquity and Early Byzantine ceramics is no longer made, and the
medieval strata are not even registered.
185
Issues regarding precise delineations of the Late Antiquity provinces have

not been considered relevant for this work.
The Fortifications of the Late Antiquity and Early Byzantine Period
139
The Province of Dalmatia A Historical Overview
With the Hunnic invasion, the majority of Illyrian towns were

destroyed.186 The decline of Roman-Byzantine towns, together with the
restricted means of artisanal industry and trade, led to these towns
being reduced to well-fortified settlements with entirely rural agglomeration. The centre shifted towards the south, to the settlements whose
crisis could be alleviated by an influx of agrarian population fleeing the
barbarians.187
The new circumstances, which emerged from the crisis of the third
century, led to the reforms of Diocletian and Constantine the Great, but
ultimately to the division of the Empire in 395. During the reign of the
Ostrogoths, Dalmatia retained the basic structure of its earlier ogranization,
but with one novelty: the merger of Dalmatia and Savia into one administrative unit that had its centre in Salona.188 Salona was an archiepiscopal
see; the existence of Dalmatian dioceses is known because of the presence
of bishops at the ecclesiastical councils in Salona in 530 and 533, which
also provided a delimitation of the province of Dalmatia.189 But it remains
unknown whether organization of dioceses was preserved after 537, when
Byzantium pushed the eastern Goths out of Dalmatia, early on in the conflict between Byzantium and the Goths. As follows from the ecclesiastical
186
187
188
189
For further information on the history of towns in the mid-400s Illyricum,

cf.: Prisci fragmenta (ed. L. Dindorf), Historici graeci minoris I, Lipsiae
1870, Frg. 2, 280.20-281.6; Frg. 8, 291.9-15; I, 7-16; . ,
, 2000, 59-60 (= , ). The
following works offer a wider account of this problem: D. S. Potter, The
Roman Empire at Bay, AD 180-395, London 2004; A. Cameron, The Later
Roman Empire, AD 284-430, Cambridge, Massachusetts 1993; S. Mitchell,
A History of the Later Roman Empire, AD 284-641: The Transformation of
the Ancient World, Hoboken, New Jersey 2007.
For further information regarding the process of disintegration and
ruralization in the hinterland of Illyricum, and the archaeological traces it
left, see: . ,
5. 7. , Sirmium (
), 2003,
239- 258; , , 58-66.
. , j , 1957, 23-24 (= ,
j ).
Diplomatiki zbornik kraljevine Hrvatske s Dalmacijom i Slavonijom I
(ured. I. Kukuljevi Sakcinski), Zagreb 1874, No. 239 and No. 240.
140
Dejan Buli
policy of Justinian I, he strived to reshape the borders of archdioceses so as

to match the borders of dioceses to those of provinces.190 Salona held its status as an archdicese, because Dalmatia was part of the Diocese of Illyricum
in the Late Empire period, with its seat in Salona.191
Bosnia and Herzegovina:

Fortifications of the Late Antiquity and the Early Byzantine Period
Bosnia and Herzegovina occupies the central part of the Balkans.

It borders with Croatia on the north, north-west and south, by the rivers
Sava and Una, and the Dinarid mountains, Serbia on the east and northeast, by the river Drina, and Montenegro on the south-east. Bosnia and
Herzegovina accesses the Adriatic Sea on the south, through the coastal
municipality of Neum.
The very name of Bosnia and Herzegovina reveals the duality of
this land. The major part of northern, peri-Pannonian Bosnia belongs to
the southern rim of the Pannonian Basin, except for the area around the
river Sava, including Semberia, which is an extension of the Pannonian
Plain. Northern Bosnia is marked by a predominantly mountainous terrain
which slopes northwards from the south.192 The mountain areas of Bosnia
and Herzegovina represent a wide expanse, part of the Dinarid mountain
range with high and medium mountains, as well as with long and deep,
often canyon-like valleys, between them. Fields of karst are by far more
numerous than basins. Eastern parts of Bosnia have karst depressions, rather
than karst fields.193
Geographically speaking, two units can be discerned in
Herzegovina: the upper or mountainous pastoral Herzegovina, and the
lower or Adriatic agricultural Herzegovina, situated in the south.194 The
mountainous Herzegovina represents the south-eastern extension of the
190
191
192
193
194
. , (
), 2004, 41-42 (= , ).
For the entire issue on the province of Dalmatia and its eastern borders, see:
, , 33-49.
, , 151-152.
, , 489-490.
. . , , 1972, 495
(= , ).
141
western Bosnian high karst, land with mountain ridges and karst fields
lying between.195 The maritime Adriatic region expands into the lower
Herzegovina, along the lower course of the river Neretva, its tributaries
and the great karst field known as Popovo polje.196
In hydrographical terms, the greatest part of Bosnia and
Herzegovina belongs to the Black Sea drainage basin, i.e. to the river Sava
basin, with the Una, Vrbas, Bosna and Drina rivers as its longest tributaries,
all flowing parallely from the south towards the north.197 A small area of
Herzegovina drains into the Adriatic Sea, with Neretva being the longest
river. Surface rivers are prevalent in northern and central Bosnia, while
subterranean rivers flow through western Bosnia and the mountainous
regions of Herzegovina.198 The lower Herzegovina is distinguished by
rivers, lost rivers, springs, surface and subterranean lakes and wetlands.
During the humid seasons of the year, karst fields become temporary lakes,
often large and deep.199
A moderate continental climate is characteristic of northern
Bosnia, while the sub-alpine climate is prevalent in the wider Dinara area.
The lower Herzegovina has the Adriatic climate, which is a variation of an
altered Mediterranean type of climate, influencing the mountainous regions
of Herzegovina as well, due to the proximity of the Adriatic coast.200
During the Late Antiquity and Early Byzantine period, the presentday Bosnia and Herzegovina approximately encompassed the hinterland of
the province of Dalmatia (Dalmatiae), as well as parts of the provinces
Pannonia Prima (Pannonia I) and Pannonia Secunda (Panonnia II).
Excavations confirmed Patschs hypothesis that castra were erected
in Doboj and ipovo (several, since castra lying on the road SalonaServitium were confirmed by the sources),201 in the aftermath of the
Pannonian uprising in the first century AD. The forms of ceramic findings
from the castrum of Doboj dated from the first to the fifth century,202 as
was confirmed by a test excavation conducted at ipovo.203 In those early
195
196
197
198
199
200
201
202
203
, , 496-497.
, , 806-807.
, , 152.
, , 490.
, , 812.
, , 490, 811.
C. Patsch, Zbirka rimskih i grkih starina u bos.-herc. Zemaljskom muzeju,
Sarajevo 1915, 57 (= Patsch, Zbirka).
I. remonik, Rimski kastrum kod Doboja, GZM 39 (1984) 70.
remonik, Rimska utvrenja, 355.
142
Dejan Buli
days, the important crossings on the tiver Sava were doubtlessly wellprotected, which in time developed into the Sava limes,204 but, not a single
fortification on the Sava has been discovered, let alone excavated, up to the
present.
Information about the movable findings are available for very few
sites, especially for the medieval ones, since published material is absent,
most of the times, despite long archaeological excavations in Bosnia and
Herzegovina, initiated at the end of the 19th century.205
Irma remonik composed a list of 263 Roman fortifications, with
emphasis on ones form the Late Antiquity. Most of these fortifications,
considering they are mainly in the highlands, were built in the Late
Antiquity or Early Byzantine period.206 But a certain number of them were
not indexed in the Lexicon: 79, 92, 93, 94, 100, 108; and some sites were
identified as prehistoric strongholds (gradine): 44, 65, 69, 70, 105; or as a
tumul: 90. Site 106 was identified as a prehistoric (gradina) and a medieval
town; site 104 as a prehistoric stronghold and a Turkish tower, while sites
28 and 30 were identified as medieval towns. We assume that in these
examples, the author probably obtained information inaccessible to us,
which led him to classify these sites as antique fortifications. But a few
sites remain problematic, as they do not appear to have been strongholds:
sites 25, 83 and 114; and it would be reasonable to omit from the list site
42 (a Roman camp deserted in the third century), site 89 (identified as a
Roman structure) and the site 203 (classified as a medieval necropolis).207
Another six sites mentioned in Perica pehars list of 60 fortifications from the Early Byzantine period,208 should be added to the list of 263
sites composed by I. remonik and incorporated into her work:209
204
205
206
207
208
209
Patsch, Zbirka, 159.

For further information regarding the history of the undertaken research,
see: Arheoloki leksikon Bosne i Hercegovine 1, Sarajevo 1988, 15-49 (=
Leksikon).
remonik, Rimska utvrenja, 355-364.
Leksikon 2-3.
The register of the fortifications, to economise space, was attached to the list
of I. remonik: remonik, Rimska utvrenja, 355-364.
pehar compiled his register without having taken into account the work
written by I. remonik: . ,
( ),
5, 2008, 17-48 (=
, ).
143
264. Gradina, Rajike, Glamo210

265. Gredine, Potoani, Livno211
266. Mareljia gradina, Staro selo-Carevica, Glamo212
267. Gradina, Prisoje-Perkovii, Duvno213
268. Gradina, Podgradina, Livno214
269. Teferi, Krupac, Ilida215
The aforementioned list should be expanded with several other
sites mentioned in the Archaeological Lexicon of Bosnia and Herzegovina.
These probably represent fortifications dating from the Late Antiquity or
Early Byzantine period and include the following:216
270. Crkvena, Kamiani, Prijedor217
271. Velika Gradina, Donja Slabinja, Bosanska Dubica218
272. Vracarevo (Vracar-grad), Brievo, Prijedor219
210
211
212
213
214
215
216
217
218
219
, , 42; Leksikon 3, 240; A. Benac, Utvrena ilirska

naselja I. Delmatske gradine na Duvanskom polju, Bukom blatu, Livanjskom
i Glamokom polju, Sarajevo 1985, 158-160 (= Benac, Ilirska naselja).
, , 42; Leksikon 3, 242; Benac, Ilirska naselja, 103-104.
, , 42; Leksikon 3, 245-246; D. Sergejevski, Putne biljeke
iz Glamoa, GZM 54 (1942) 153; Benac, Ilirska naselja, 180-181.
, , 42; Leksikon 3, 266; V. Radimsky, Starine kotarska
upanjog u Bosni, GZM 6 (1894) 300; Benac, Ilirska naselja, 21.
M. Mandi, Gradine, gromile i druge starine u okolini Livna, GZM 47 (1935)
12; , 42; Leksikon 3, 239.
, , 42; Leksikon 3, 57; . ,
, ,
2003, 103; D. Sergejevski, Arheoloki nalazi u
Sarajevu i okolici, GZM 2 (1947) 46-48.
The deficiencies of this classification are evident; a considerable number of
these forts were categorized only after surface findings, collected during
reconnaissance. Scarce information from the Lexicon often omit pottery
findings, while the chronological classification is most often given with a
simple, broad phrase Late Antique fortification.
Leksikon 2, 34.
Leksikon 2, 39.
V. Radimsky, O nekojim prehistorijskim i rimskim graevnim ostacima u
podruju rijeke Sane u Bosni, GZM 3 (1891) 439-440; D. Sergejevski,
Epigrafski nalazi iz Bosne, GZM 12 (1957) 112-116; D. Sergejevski, Rimski
rudnici eljeza u sjeverozapadnoj Bosni, GZM 18 (1963) 88-92; Leksikon
2, 39.
144
Dejan Buli
273. Mali Grad-Blagaj kod Mostara.220

274. Cetinac, Bokovii, Laktai221
275. Lisiji Brijeg-Cintor, Laminci, Bosanska Gradika222
276. Gradac, Trnovica, Zvornik223
277. Veliki Gradac, Ostojievo, Bijeljina224
278. Zvornik 1, Zvornik225
279. Crkvena, Dragoaj, Banja Luka226
280. Gradina, Brdo-Rudii, Mrkonji Grad227
281. Gradina, Bosansko Grahovo228
282. Gradina, Drvar Selo-Glavica, Drvar229
283. Velika Gradina, Lastve-Rakovice, Bosanski Petrovac230
284. arampovo, Gornji Vakuf231
285. Babunar (Saraj), Travnik232
286. Blace, Rankovii, Pucarevo233
220
221
222
223
224
225
226
227
228
229
230
231
232
233
. Basler, Arhitektura kasnoantikog doba u Bosni i Hercegovini, Sarajevo

1972, 50 (= Basler, Arhitektura).
Leksikon 2, 48.
L. eravica - Z. eravica, Arheoloka nalazita u okolini Bosanske Gradike,
Zbornik Krajikih muzeja 6, Banja Luka 1974, 220-221; Leksikon 2, 52.
Leksikon 2, 91.
C. Patsch, Mali rimski nahoaji i posmatranja, GZM 9 (1897) 518; Leksikon
2, 98.
M. Babi, custodian of the museum in Bijeljina, has confirmed the existence
of an Early Byzantine layer by means of sondage, of which he was kind to
let us know. . Mazali, Zvornik (Zvonik). Stari grad na Drini, GZM Istorija
i etnografija 10 (1955) 73-116; D. Kovaevi-Koji, Zvornik (Zvonik) u
srednjem vijeku, Godinjak drutva istoriara BiH 16, Sarajevo 1967, 19-35;
Leksikon 2, 98.
Leksikon 2, 128.
Leksikon 2, 146.
I. remonik, Dva srednjovekovna grada u okolici Grahova, GZM 8 (1953)
349-351; Leksikon 2, 161.
V. uri, Starine iz okoline Bosanskog Petrovca, GZM 14 (1902) 252; Z. Vinski,
Kasnoantiki starosjedioci u salonitskoj regiji prema arheolokoj ostavtini
predslavenskog supstrata, VAHD 69, 1967 (1974) 41; Leksikon 2, 162.
V. uri, Starine iz okoline Bosanskog Petrovca, GZM 14 (1902) 22-23;
Leksikon 2, 165-166.
J. Petrovi, Novi arheoloki nalazi iz doline Gornjeg Vrbasa, GZM 15-16
(1960-1961) 1961, 231-234; Basler, Arhitektura, 84; Leksikon 2, 186.
P. A. Hoffer, Nalazita rimskih starina u travnikom kotaru, GZM 7 (1895) 50
(= Hoffer, Nalazita); J. Koroec, Travnik i okolina u predhistorijsko doba,
GZM 4-5 (1949-1950) 1950, 254-265 (= Koroec, Travnik); Leksikon 2, 195.
Leksikon 2, 195.
287. Glavica, Mali Mounj, Vitez234

288. Gradac (Tarabovac), Vilenica, Travnik235
289. Gradina-Megara, Gole, Travnik236
290. Grbavica Brdo, Grbavica, Vitez237
291. Jankovii, Jankovii, Travnik238
292. Oblak, Mali Mounj-Divljaci, Vitez239
293. Trojan, Pazari, Hadii240
294. Domavia, Gradina-Sase, Srebrenica241
295. Rade, Neum. Sitomir, Radiii, Ljubuki242
296. Veliki vrh, Romanija, Sokolac243
297. Veliki Gradac, Presjeka-Mahinii, Nevesinje244
298. Brijeg, Pareani, Bilea245
299. Gradina, Brova, Trebinje246
234
235
236
237
238
239
240
241
242
243
244
245
246
Koroec, Travnik, 257; Leksikon 2, 197.

Hoffer, Nalazita, 54; Koroec, Travnik, 259; Leksikon 2, 199.
Koroec, Travnik, 258; D. Sergejevski, Novi i revidirani rimski natpisi, GZM
6 (1951) 309; Leksikon 2, 203.
Leksikon 3, 57.
L. Pogatschnig, Stari rudokopi u Bosni, GZM 2 (1890) 125-130; V.
Radimsky, Rimski grad Domavija u Gradini kod Srebrenice u Bosni i
tamonji iskopi, GZM 3 (1891) 1-19; F. Buli, Rimski nadpisi u Srebrenici
(Municipium Domavia), GZM 3 (1891) 387-390; V. Radimsky, Prekopavanje
u Domaviji kod Srebrenice godine 1891., GZM 4 (1892) 1-24, C. Patsch,
Prilozi naoj rimskoj povjesti, GZM 22 (1910) 1911, 192-195; D. Sergejevski,
Epigrafski i arheoloki nalazi (ipovo, Livno, Duvno), GZM 42, sv. 2 (1930)
162-163; D. Sergejevski, Rimski natpisi iz Bosne, uikog kraja i Sandaka,
Spomenik SKA 93, Beograd 1940, 144; I. Bojanovski, Biljeke iz arheologije
I, Nae Starine 19 (1964) 193; I. Bojanovski, Arheoloki pabirci sa podruja
antike Domavie. lanci i graa za kulturnu istoriju istone Bosne, Tuzla
1965, 103; Leksikon 3, 69.
C. Patsch, Mali rimski nahoaji i posmatranja, GZM 9 (1897) 528-529;
Leksikon 3, 334.
. Truhelka, Prethistorijske gradine na Glasincu, GZM 3 (1891) 306-307;
Leksikon 3, 108.
D. Sergejevski, Rimska cesta na nevesinjskom polju, GZM 3 (1948) 55; I.
Bojanovski, Prilozi za topografiju rimskih i predrimskih komunikacija i
naselja u rimskoj provinciji Dalmaciji (s posebnim obzirom na podruje
Bosne i Hercegovine). II - Prethistorijska i rimska cesta Narona - Sarajevsko
polje s limotrofnim naseljima, Godinjak Akademije nauka i umetnosti
Bosne i Hercegovine 17, Sarajevo 1978, 90-91; Leksikon 3, 153.
Leksikon 3, 170.
Leksikon 3, 177.
145
146
Dejan Buli
300. Velika Gradina, Slivnica, Trebinje247

301. Vraevica, Panik, Bilea248
302. Grad Lis, Repovci, Konjic249
303. Gradac, Glavatievo, Konjic250
304. Ilina, Gorani, Konjic251
305. Velika Gradina, Varvara, Prozor252
306. Anelia (Juria) Gradina, Lipa, Livno253
307. Gradina, Podgradina, Livno254
308. Gradina (Nuhbegovia gradina), Podhum, Livno255
309. Kasalov Gradac, Livno256
310. Brina, Vinjani, Posuje257
311. Bukovac 2, itluk, Posuje258
312. Grad, Stipanii, Duvno259
313. Gradina, Korita, Duvno260
247
248
249
250
251
252
253
254
255
256
257
258
259
260
. Odavi, Praistorijska nalazita na prostoru Trebinja (gomile i gradine),

Tribunia 4, Trebinje 1978, 153; Leksikon 3, 195.
I. Bojanovski , Arheoloki spomenici, Nae starine 8 (1962) 12; Leksikon 3, 196.
P. Aneli, Historijski spomenici Konjica i okoline, Konjic 1975, 158-160 (=
Aneli, Historijski spomenici); Leksikon 3, 213.
P. Aneli, Srednjovekovni gradovi u Neretvi, GZM 13 (1958) 200-202;
Leksikon 3, 213.
Aneli, Historijski spomenici, 29; Leksikon 3, 217.
V. uri, Gradina na vrelu Rame, prozorskog kotara, GZM 12 (1900) 99118; . Truhelka, Kulturne prilike Bosne i Hercegovine u doba prethistorije,
GZM 26 (1914) 79-80; B. ovi, Prelazna zona, Praistorija Jugoslovenskih
zemalja 4 (1983) 390-412; N. Mileti, Rani srednji vijek, Kulturna istorija
Bosne i Hercegovine od najstarijih vremena do pada ovih zemalja pod
osmansku vlast, Sarajevo 1984, 422; Leksikon 3, 225.
V. uri, Arheoloke biljeke iz Livanjskog kotara, GZM 21 (1909) 169170; M. Mandi, Gradine, gromile i druge starine u okolini Livna, GZM 47
(1935) 9-10; A. Benac, Utvrena ilirska naselja, I. Delmatske gradine na
Duvanskom polju, Bukom blatu, Livanjskom i Glamokom polju, Sarajevo
1985, 134 (= Benac, Ilirska naselja); Leksikon 3, 235.
Leksikon 3, 239.
V. uri, Arheoloke biljeke iz Livanjskog kotara, GZM 21 (1909) 169;
Benac, Ilirska naselja, 80-83; Leksikon 3, 239-240.
M. Mandi, Gradine, gromile i druge starine u okolini Livna, GZM 47 (1935)
7; Benac, Ilirska naselja, 108-110; Leksikon 3, 244.
Leksikon 3, 260.
P. Ore, Prapovjesna naselja i grobne gromile, GZM 32 (1977) 1978, 218-219 (=
Ore, Naselja); Leksikon 3, 261.
Leksikon 3, 264.
N. Mileti, Ranosrednjovekovna nekropola u Koritima kod Duvna, GZM 33
(1978) 1979, 141-204, . Miki, Rezultati antropolokih ispitivanja ranosre-
147
314. Vukove Njive, Gradac, Posuje261

315. Bilobrig, Vionica, itluk262
316. Gradina, Mali Ograenik-Donji Ograenik, itluk263
317. Krstina, Hamzii, itluk264
318. Mala Gradina, apljina265
319. Milanovaa, Gorica, Grude266
320. Trebinje-Crkvine267
With these additions, we reach a total of 320 fortifications, mainly
from the Late Antiquity or Early Byzantine period. This figure still does
not reflect their real quantity, with all the already mentioned deficiencies
of such a classification and some zones having been poorly explored, but it
is certainly closer to the actual number. The empty zones were not uninhabited in the Late Antiquity, for these were the mining districts of eastern Bosnia or the fertile valleys around the Bosna river. A lot of strongholds (gradine) were, with inertia, were dated of as prehistoric. But even
if we accept such datings, there remains a number of Late Medieval towns
whose Late Antiquity or Early Byzantine phase can be assumed to exist.
The conjectured density of fortifications can be glimpsed at by comparing
the empty zones with the surrounding ones.
Since the historical information being absent and the adequate
archaeological information being scarce, it is difficult to speak of the historical context beyond general observations. The process of adapting to the
new circumstances unfolded in two directions. The first was fortifying the
already existing settlements in the plains, as seen in Mogorjelo at apljina,
where an agricultural estate was fortified already in the early fourth century.
The other direction, far more efficient, is the so-called vertical migration
resettlement to the fortifications on higher altitudes.268
261
262
263
264
265
266
267
268
dnjovekovne nekropole u Koritima kod Duvna, GZM 33 (1978) 1979 205222; Benac, Ilirska naselja, 74-76; Leksikon 3, 264-265.
Ore, Naselja, 184-185; Leksikon 3, 279.
Leksikon 3, 290.
. Belagi, Steci. Kataloko topografski pregled, Sarajevo 1971, 315 (=
Belagi, Steci); Leksikon 3, 297.
Leksikon 3, 301.
C. Patsch, Pseudo-Skylaxovo jezero. Prinos povjesti donjeg poreja Neretve,
GZM 18 (1906) 374-376 (= Patsch, Pseudo-Skylaxovo); Leksikon 3, 330.
Patsch, Pseudo-Skylaxovo, 379; Leksikon 3, 331.
. , 7. 10. , 2007, 158 (=
, ).
, , 37.
148
Dejan Buli
But this does not exclude the possibility that the exploitation of fertile plains, suitable for agricultural production, could have continued. We
can speak of a more large-scale fortification construction in the hinterland
of Dalmatia only after 535 and the final expulsion of the Goths from
Dalmatia, since it is unlikely that during their reign they would allow living
in strongholds.269 Besides, the number of the known fortifications in continental Croatia is, so far, meagre.
Perica pehar divided the fortifications in four big groups, based on
a sample of 60 fortifications from Late Antiquity or Early Byzantine period,
according to their surface area: big, middle-sized and small, while the fortifications with an unknown surface made a group of its own.270 Small fortifications, in the hinterland of Dalmatia, represent the most numerous group.
As ihajlo ilinkovi warned, when classifying the fortifications
according to their size, one should be aware that, most of the times, the
outer extensive ramparts often remained undiscovered, and that they
could have been used occasionally to keep the livestock during the siege.271
pehars division may be accepted, but it should be borne in mind
that all the fortifications on high altitudes were located on more or less
steep slopes. When making a projection of a ground plan, which is normally executed on a horizontal plane, shrinking of the surface area unavoidably happens, in line with the laws of mathematics.272 But the conclusions
that the big-sized fortifications, erected on the elevations overlooking the
fertile plains, rivers or fields, acted as a sort of collective centres in addition to having a defensive role, and maybe even that of ore storagesremain dubious.273 One of the main functions the fortifications had was
probably the protection of the mining basins and auriferous rivers.
269
270
271
272
273
Procopius makes no mention of fortification construction in Dalmatia.

The first group is made up of fortifications with a surface area greater than
1 hectare; the second of fortifications with a surface area between 0.5 and 1
hectare; while the fortifications of a surface area smaller than 0.5 hectare
fall into the third group: , , 19.
. , . , 2010, 225-226.
In order to take the measurements of the surface area, it is necessary to have
in mind the shrinkage that occurs when projecting terrain onto a flat
horizontal plane, except for where there are no slopes and the surface
remains the same. Practically, this would mean that the represented surface
of the fortification is 86.6% of the real one, if the angle of the slope is 30;
and only 70.7%, if the angle of the slope is 45. It is an entirely different
question if some surfaces are useful due to these terrain slopes.
, , 38.
149
Findings of slag indicate that the fortifications were erected in the vicinity of the mining shafts, and the residues of slag are frequently found on
many sites, regardless of their geographical position or size, as had been
suggested. The idea that the discovered buildings had the function of
horeum (silo for storage of agricultural products) has no foundation.
Positions these fortifications occupied could determine their main
tasks and functions; however, the excavations carried out in or around
these sites so far do not yield sufficient elements that could make a correlation between the surface of a fortification and its function. The crucial
function of the fortifications situated along the main roads was to secure
the traffic, settlements or river crossings. Besides the insufficient research
on the fortifications and the deficient knowledge of the traffic ways (especially the less significant ones), additional difficulty lies in the locations of
a majority of Roman settlements that we know of from the sources,
remaining unidentified.274
On the other hand, perceived clusters of fortified points along the
border of the maritime Adriatic belt and on the mountain massifs that
separated the coastal regions from the hinterland of Dalmatia are spurious
as well.275 We think that such attitude comes, doubtlessly, from the insufficient research of the given areas that led to the false clusterization of the
fortified points. Also, without understanding that these generally represented fortified villages,276 with no military function, this theory should be
rejected. Nevertheless, the unquestionably higher density of fortifications
274
275
276
remonik, Rimska utvrenja, 357.

Other than the local functions - protection of roads and settlements - the
fortifications around Bosanski Petrovac, Grahovo, Livanjsko polje,
Glamoko polje, Duvanjsko polje, Posuje, Gruda, Imotsko polje, Ljubuko
polje, and those lying along the lower course of Neretva, formed a solid
barrier towards the hinterland; See: remonik, Rimska utvrenja, 357.
In the last couple of years, an opinion prevailed that most of the
fortifications served as fortified settlements, without excluding additional
functions. The nature of the archaeological findings confirms this
hypothesis, since these have been predominantly associated with
craftsmanship and agriculture, and there are objects pointing to the
presence of women and existance of churches, all indicating a longer stay
within the forts. Cf: . ,
,
, - 2002, 7172; . ,
, III, 2005, 180; . ,
. ,
2010, 227.
150
Dejan Buli
comes as a consequence of geographic conditions i.e. the fact that these

were erected on a low, coastal stretch of land which led some inhabitants
to leave the area for the island fortifications, and the majority to flee to the
highlands of the Dinara mountains. Most likely such process of receding
was happening on the northern side of the massif as well.
The following, revised list, includes the fortifications that, besides
the already mentioned Late Antique/Early Byzantine strata, contain medieval
traces that indicate a continuous or re-initiated use of the fortification. 277
1. Brekovica, Biha (95)278
2. Zecovi, arakovo, Prijedor (81)279
3. Grad, Gornji Vrbljani, Klju (Velika and Mala Gradina (80)280
4. Gradina (Grad), Gradac, Posuje (46)281
5. Zelengrad, Han Kola-utkovci, Banjaluka (134)282
277
278
279
280
281
282
The number within the parentheses designates the number of the site,
corresponding to the number on the provided map.
Leksikon 2, 14. Some authors date the remains of ramparts and of the
pentagonal tower only to the Late Antiquity and the Early Byzantine
period: V. Radimsky, Nekropola na Jezerinama u Pritoci kod Bia, GZM 5
(1893) 41; P. pehar, Late Antique and Early Byzantine Fortification in
Bosnia and Herzegovina (Hinterland of the Province of Dalmatia),
Hhensiedlungen zwischen Antike und Mittelalter-Ergnzungsbnde zum
Reallexikon der Germanischen Altertumskunde, Band 58, Berlin New
York 2008, 586 (= pehar, Late Antique).
The foundations of the church, as well as the sporadic medieval objects
confirm that these fortifications were used in the Middle Ages: Leksikon 2,
39; I. remonik, Rimski ostaci na Gradini Zecovi, GZM 11 (1956) 137-146;
Basler, Arhitektura, 55.
The occupation continued into the Carolingian age (8th 9th century). That
is confirmed by the archaeological findings such as the ceramics of Early
Slavonic type, a bronze spur and a gold-plated prong of a belt buckle:
Leksikon 2, 144; Z. Vinski, Novi ranokarolinki nalazi u Jugoslaviji, Vjesnik
Arheolokog muzeja u Zagrebu 10-11 (1977-78) 1979, 143-190; I.
Bojanovski, Kasnoantiki katel u Gornjim Vrbljanima na Sani, GZM 34
(1979) 1980, 109-119.
In some of the researched structures on the slopes of gradina were noticed
material remains of the Early Medieval period (the Slavic period): Leksikon
3, 264.
Remnants of the wall above the Late Antique fortification are thought to be
related to the town of Zemljanik, mentioned in the sources from the late
13th century: . , ,
GZM 48 (1936) 33. West of the plateau, a necropolis arranged in rows was
discovered and categorized as medieval: Leksikon 2, 133.
6. Mogorjelo, apljina (252)283

7. Biograci, Litice, Mostar (37)284
8. Gradac, Hudutsko, Prozor (29)285
9. Gradina, Bivolje brdo, apljina (263)286
10. Grad Biograd, Zabre, Konjic (24)287
11. Blagaj (Stjepan Grad), Blagaj, Mostar (35)288
12. Vidoki Grad, Stolac (191)289
13. Gradina, Alihode, Travnik (68)290
14. Crkvina, Makljenovac, Doboj (73)291
283
284
285
286
287
288
289
290
291
Besides the necropolis dating from the Late Antiquity and the Middle Ages,
on the fortified site and in its immediate surroundings human habitation in
the Early Middle Ages was confirmed with the medieval ceramics and Early
Carolingian findings. Several tombstones (steci) have also been preserved,
so the archaological findings cover the period from the eighth to fifteenth
century; . Werner, Ranokarolinka pojasna garnitura iz Mogorjela kod
apljine (Hercegovina), GZM 25-26 (1961) 235-242; Z. Vinski, O nalazima
karolinkih maeva u Jugoslaviji, SP 11 (1981) 9-54; Z. Vinski, Zur
karolingischen Schwertfunder aus Jugoslawien, Jahrbuch des RmischGermanischen Zentralmuseums Mainz 30 (1983) 465-501; Leksikon 3, 331.
This fortification was again used during the eighth and ninth centuries; by
a population of Slavic characteristics, but under Frankish influence, as
confirmed by the discovered spur: I. remonik, Rimsko utvrenje na
Gradini u Biogracima kod Litice, GZM 42/43 (1989) 89-92.
The use of Gradac in the Middle Ages has been confirmed by the findings
of Late Medieval ceramics: Leksikon 3, 213.
Discovered movable findings represent pre-historic and Roman ceramics
and bricks, so it remains unclear why a medieval settlement was even
mentioned: Leksikon 3, 325.
This site was mentioned in 1444, 1448 and 1454 as the domain of Herzeg
Stjepan. In the Turkish census of 1469, it was mentioned as a deserted town,
while the square (trg, a suburb) of the same name had 17 houses: Leksikon
3, 213; P. Aneli, Historijski spomenici Konjica i okoline, Konjic 1975,
125-129 (= Aneli, Historijski spomenici).
The earliest source that explicitly mention the town dates back to 1423. The
Turks took the town in 1465 and in the eighteenth century the walls of this
structure were once again redesigned. What particularly draws attention is
a twelfth-century stone plate with a cyrillic inscription, in a secondary use:
This town was mentioned for the first time in the fifteenth century and it
was destroyed later, during the construction of Austro-Hungarian barracks:
Leksikon 3, 195; Basler, Arhitektura, 50-51.
A fragment of Early Medieval (Slavic) ceramics was discovered in the area
of Gradina: Leksikon 2, 198.
During the Middle Ages, there was a wooden church on the hilltop with graves
around it dated from the ninth to thirteenth centuries: Leksikon 2, 63.
151
152
Dejan Buli
15. Bobovac, Dragovii-Miljakovii, Vare (63)292

16. Gradac, Homolj, Kiseljak (59)293
17. Gradina, Dabravina, Vare (171)294
18. Teferi, Krupac, Ilida (269)295
19. Crkvena, Kamiani, Prijedor (270)296
20. Bosanska Gradika, Bosanska Gradika (113)297
21. Lisiji Brijeg-Cintor, Laminci, Bosanska Gradika (275)298
22. Zvornik 1, Zvornik (278)299
23. Gradina, Bosansko Grahovo (281)300
292
293
294
295
296
297
298
299
300
For the first time Bobovac was mentioned in 1350, while a royal court was
being built from the second half of the fourteenth to the mid-fifteenth
century. The Turks took it in 1463: P. Aneli, Bobovac i Kraljeva Sutjeska.
Stolna mesta bosanskih vladara u XIV i XV stoleu, Sarajevo 1973; Leksikon
3, 15. For further information regarding remains from the Early Byzantine
period, see: . Basler, Kanelirani stup iz Stoca, Slovo Gorina 10, 1982, 52-53.
Besides one medieval ceramic vessel, graves dated to the Middle Ages were
discovered above the Early Byzantine basilica: Leksikon 3, 19. pehar claims
that these tombs have to be dated to the Late Antiquity: pehar, Late
Antique, 573. V. Skari, Altertmer von Gradac in der Lepenica (Bosnien)
(Starine na Gracu u bosanskoj Lepenici), GZM 44 (1932) 1-21.
Individual medieval findings were found inside the Gradina. These include
several objects made of iron and a trefoil arrow, dated to the Early Middle
Ages. The issue of dating these objects to the Antiquity or the Middle Ages
remains open: D. Sergejevski, Bazilika u Dabravini (Revizija), Sarajevo 1956;
I. Nikolajevi, Kasnoantike presvoene grobnice u srednjovekovnoj
crkvenoj arhitekturi Bosne i Hercegovine, Predslavenski etniki elementi na
Balkanu u etnogenezi Junih Slovena, Sarajevo 1969, 223-227. I. Nikolajevi,
Oltarna pregrada u Dabravini, ZRVI 12 (1970) 91-112; For a more
generalized overview, see: Leksikon 3, 19.
D. Sergejevski and K. Topolovac claim that this was a late medieval fortification: D. Sergejevski, Arheoloki nalazi u Sarajevu i okolini, GZM 2, (1947)
46; Leksikon 3, 57, while M. Popovi and P. pehar support the theory of
Late Antique/Early Byzantine fortification: , ,
103; pehar, Late Antique, 586.
Leksikon 2, 34.
E. Paali, Antika naselja i komunikacije u Bosni i Hercegovini, Sarajevo
1960, 27; L. eravica - Z. eravica, Arheoloka nalazita u okolini Bosanske
Gradike, Zbornik Krajikih muzeja 6, Banja Luka 1974, 215-233 (= eravica
- eravica, Arheoloka nalazita); G. Kraljevi, Rimski novci iz Bosanske
Gradike i Laktaa, GZM 34 (1978) 1979, 137.
eravica - eravica, Arheoloka nalazita, 220-221; Leksikon 2, 52.
Leksikon 2, 98; . Mazali, Zvornik (Zvonik). Stari grad na Drini, GZM
Istorija i etnografija 10 (1955) 73-116; D. Kovaevi-Koji, Zvornik (Zvonik)
u srednjem vijeku, Godinjak drutva istoriara 16, 1967, 19-35.
I. remonik, Dva srednjovekovna grada u okolici Grahova, GZM 8 (1953)
349-351; Leksikon 2, 161.
24. Glavica, Mali Mounj, Vitez (287)301

25. Gradina-Megara, Gole, Travnik (289)302
26. Kastel- Banja Luka (76)303
27. Veliki vrh, Romanija, Sokolac (296)304
28. Grad Lis, Repovci, Konjic (302)305
29. Gradac, Glavatievo, Konjic (303)306
30. Velika Gradina, Varvara, Prozor (305)307
31. Gradina (Nuhbegovia gradina), Podhum, Livno (308)308
32. Gradina, Korita, Duvno (313)309
33. Vukove Njive, Gradac, Posuje (314)310
34. Gradina, Mali Ograenik-Donji Ograenik, itluk (316)311
301
302
303
304
305
306
307
308
309
310
311

Koroec, Travnik, 257; Leksikon 2, 197; Belagi, Steci, 145; Leksikon 2, 199.
In the thirteenth century, Banja Luka belonged to the upa Zemljanik and
the oblast (area) of Donji Kraji. Its modern name was mentioned for the first
time in 1494. After the fall of the Bosnian state (1463), Banja Luka became
a part of the banovina of Jajac, and the Turks took it in early 1528: A. Bejti,
Banja Luka pod turskom vladavinom, Nae Starine 1 (1953) 91-116; V.
Skari, Banja Luka i njena okolina u davnini, Otabina 31-33 (1924), 2;3;2;
I. remonik, Kastel Banja Luka. Gradina sa slojevima od praistorije do
danas, AP 14 (1972) 133-134; L. eravica, Kastel Banja Luka. Kompleksno
utvrenje, AP 15 (1973) 112-113; B. Graljuk, Posavina u antici u svjetlu
novih istraivanja, Antiki gradovi i naselja u junoj Panoniji i graninim
podrujima, Varadin 1977, 147-154; Banja Luka, Enciklopedija Jugoslavije
1, A-Biz, Zagreb 1980, 492-494 (M. Vasi); Leksikon 2, 130; D. Peria,
Zlatnik cara Justinijana iz Banjaluke, GZM 45 (1990) 171-176.
Leksikon 3, 108.
Aneli, Historijski spomenici, 158-160; Leksikon 3, 213.
P. Aneli, Srednjovekovni gradovi u Neretvi, GZM 13 (1958), 200-202;
Leksikon 3, 213.
V. uri, Gradina na vrelu Rame, prozorskog kotara, GZM 12 (1900) 99118; N. Mileti, Rani srednji vijek, Kulturna istorija Bosne i Hercegovine od
najstarijih vremena do pada ovih zemalja pod osmansku vlast, Sarajevo
1984, 422; Leksikon 3, 225.
V. uri, Arheoloke biljeke iz Livanjskog kotara, GZM 21 (1909) 169; A.
Benac, Utvrena ilirska naselja, I. Delmatske gradine na Duvanskom polju,
Bukom blatu, Livanjskom i Glamokom polju, Sarajevo 1985, 80-83;
N. Mileti, Ranosrednjovekovna nekropola u Koritima kod Duvna, GZM 33
(1978) 1979, 141-204; . Miki, Rezultati antropolokih ispitivanja
ranosrednjovekovne nekropole u Koritima kod Duvna, GZM 33 (1978)
1979, 205-222; Benac, Ilirska naselja, 74-76; Leksikon 3, 264-265.
Leksikon 3, 279.
Leksikon 3, 297.
153
154
Dejan Buli
35. Mali Grad-Blagaj near Mostar (273)312

36. Grad Vitanj, Kula, Sokolac (106)313
37. Gradina Loznik, Podloznik, Pale (104)314
38. Gradina Bokaevac, Kostajnica, Konjic (28)315
39. Gradina, Vraba, Bijela, Konjic (30)316
40. Vrtine, rvanj, Ljubinje (203)317
41. Trebinje-Crkvine (320)318
This leads us to the conclusion that out of 320 Late Antique/Early
Byzantine sites, medieval traces appear on 41 sites, or 12.81%. We hold
this percentage to be much higher in reality, which can be deduced if we bear
in mind the deficiencies and scarcity of information, because of which
medieval horizons are impossible to discern.
And since the sites taken into account here were often merely
registered in the process of reconnaissance, or yielded only scarce and
inaccurately dated findings, a wider picture and chronological frame of
these sites has proved very complex to grasp. The absence of written
sources and infrequent occurrence of the remaining architectural
monuments add to the complexity of this task, as well.
A more accurate dating of certain fortifications has not been
established beyond them being medieval towns: 1, 23, 31, 36, 38, 39; some
represented a medieval town with a church in it: 2; or a medieval town and
a necropolis: 5. When it comes to site 18, only a broad conclusion can be
made that it belongs to the Middle Ages. Sites 20 and 21 were classified as
312
313
314
315
316
317
318
About 2.5 km from the fortification of Blagaj near Mostar, stands Mali
Grad, formed of a tower, what was probably a cistern, and another
building. The ground floor of the tower corresponds with the time of
Emperor Justinian I: Basler, Arhitektura, 50.
remonik, Rimska utvrenja, 360; Leksikon 3, 96.
Belagi, Steci, 263; remonik, Rimska utvrenja, 360; Leksikon 3, 54.
remonik, Rimska utvrenja, 358; Aneli, Historijski spomenici, 163-167;
250-255; Leksikon 3, 215.
remonik, Rimska utvrenja, 358; P. Aneli, Srednjovekovni gradovi u
Neretvi, GZM 13 (1958) 185-189; Aneli, Historijski spomenici, 129-133;
Leksikon 3, 215.
Belagi, Steci, 379; remonik, Rimska utvrenja, 361; Leksikon 3, 196.
Archaeological excavations confirm existence of a town, about 1.2 ha in
surface area. Accidental pottery findings point to the Early Byzantine
period the seventh century, as well as to the period between the ninth and
tenth centuries: . , 7. 10. ,
2007, 158 (= , ).
155
medieval settlements; site 19 as a medieval building, while individual

medieval findings were discovered on several sites: 16, 17, 24. Slightly
more precise designations were provided for sites 8, 10, 12, 15, 22, 26, 28?,
29? as Late Medieval towns; site 33 was classified as a Late Medieval settlement; the following sites were identified as Late Medieval necropoles:
32, 34, 40; site 25 as a tombstone; site 11 was dated to the Late Medieval,
Ottoman period; site 37 was identified as a Turkish tower. Site 14 was a
indentified as church with a necropolis, dated between the 9th and 13th
centuries.
According to Slavic and Early Carolingian findings, the following
sites were defined as Early Medieval: 3, 4, 6, 7, 13; site 30, which represent a
settlement with a necropolis, was also dated to the Early Medieval period.
Site 9 was dated to the Middle Ages, for unknown criteria; site 27 was
destroyed during later construction works, which might corroborate the
hypothesis that it dates back to the Middle Ages.
Years after the fall of Salona represented the beginning of a new
age, one of continuous Slavic settlement in the decades that followed. In
the second wave of migrations, with the emperors consent, the Serbs and
the Croats got hold of the entire area of the former province of Dalmatia,
where the first principalities would rise some time later. About them we
know from the treatises of Emperor Constantine VII Porphyrogenitos.319
Byzantine coastal towns and some islands were the only ones spared of the
conquests and they will play an important role in Christianization and the
development of Slavic hinterland.320
There is almost no historical information on the events in Bosnia
and Herzegovina during the first couple of centuries after the Slavic colonization, and the archaeological insights hardly provide a more profound
perspective. Opportunities were not taken adequately, just because many
sites with these remains were either excavated too early at the turn of
the century, or too late destroyed before being researched.
Why the architectural elements attributed to the Slavs are difficult
to recognize will be discussed later; for now, it will suffice to acknowledge
their presence in the strongholds (gradine). During excavation of the
319
320
Constantine Porphyrogenitus, De administrando imperio I (ed. Gy.

Moravcsik R. J. H. Jenkins), Washington DC 1967, cc 31-36 (= DAI).
, , 309.
156
Dejan Buli
fortifications, fragments of early Slavic ceramics were discovered. These

findings reflect the attitude of the Slavs towards their new environment,
but the use of these sites is not an evidence for the adaptation of the newcomers to the earlier settlements, nor is it a proof for the continuity of life.
However, it is a proof of analogous factors that led to the fortifications
being re-used immediate war danger, in this case. Purposely chosen and
situated on important strategic points, they justified the reason of their
choice and affirmed their centuries-long importance.
The first to mention Bosnia was Constantine Porphyrogenitos in
the mid-tenth century, when it was still a part of Serbia, while other lands
lying within the province of Dalmatia were principalities of the Narentines,
Zachlumia and Travunia, ruled by archonts. Salines (in the vicinity of the
present-day Tuzla) was included as well, among other Serbian towns,
whereas only two towns in Bosnia were mentioned, Katera and Desnik.321
Katera was thought to be Kotorac near Sarajevo, but this site has no
medieval strata whatsoever; it could have been Kotor, in the middle of the
Vrbanje upa (administrative unit). It has been known under the name of
Bobac (Bobos), but all that is known of the town pertains to the Late
Middle Ages. The location of Desnik remains unidentified, but it was
thought to be located near the present-day Deanj.322 Alternatively, if we
follow the understanding that the term kastra oikoumena in De administrando imperio, the treatise of Constantine VII Porphyrogenitos, does not
designate inhabited towns, but lists the towns in the ecclesiastical organization of the Roman church, these two towns might be Bistua (Zenica or
Vitez) and Martar (Mostar or Konjic).323
Porphyrogenitos mentions five towns in Travunia: Trebinje, Vrm,
Risan, Lukavete and Zetlivi;324 in Zachlumia aside from Bona and Hum,
another five: Ston, Mokriskik (Mokro), Josli (Olje), Galumajnik and
Dobriskik;325 and among the Narentines (Pagans) the towns of Rastoka and
Dalen (Doljani). Risan is a well-known coastal town in Montenegro. Trebinje
was founded at the site of the present-day Crkvine, over an earlier Roman
fortification. Accidental findings of pottery were dated to the Early
321
322
323
324
325
DAI I, 32.149-151.
,
2010, 183.
T. ivkovi, On the Beginnings of Bosnia in the Middle Ages, Spomenica
akademika Marka unjia (1927-1998), Sarajevo 2010, 177-178.
DAI I, 34.19-20.
DAI I, 33.20-21.
157
Byzantine period and the 7th century, and fragments from the 9th - 10th centuries were found next to the ramparts.326 The position of Vrm has not been
established yet, but it is being searched for around the Trebinjica river
east of Trebinje (maybe around Panik). Lukavetija and Zetlivija have not
been localized with certainty.327
Bona and Hum were, in all likelihood, located at the site of Blagaj
beside Mostar. Smaller forts were erected on two hilltops, Stjepan grad and
Mala gradina, outside which settlements existed probably already in the
Early Middle Ages, which corresponds to the reports by Constantine
Porphyrogenitos on these two towns.328
In the tenth century, Bosnia was a part of the Serbian realm, ruled
by prince aslav. And it seems that after his death, in the mid-tenth century, Bosnia broke off and became politically independent.329 At the close
of the century, it was subjugated by the Bulgarian tsar Samuil, and afterwards became a part of the Byzantine Empire. Throughout the 11th century,
Bosnia, Travunia and Zachumlie were under the authority of the Doclean
state. From the mid-twelfth century, Bosnia was under the supreme rule
of Hungary, followed by a brief return to Byzantium. Then began a new
age for Bosnia and Herzegovina that would last until the Ottoman conquest
of Bosnia in 1463, and of Herzegovina in 1481.330
In all these times of war, the fortifications were more or less used,
but as no systematic excavations took place until today, it is guesswork to
say when and under what circumstances were some of them sites of war
operations, which are proven by remains of weapons and traces of fire on
some of the sites.
326
327
328
329
330
, , 158.
For further information regarding the proposed ubications, see: . ,
II ( ).
, 48 (1880) 1-152; .
, , 37
(1998) 20-21; . Loma, Serbischen und kroatisches Sprachgut bei Konstantin
Porphyrogennetos, 38 (1999/2000) 87-160; T. , Constantine
Porphyrogenitus Kastra oikoumena in the Southern Slavs Principalities,
57 (2008) 9-28 .
Basler, Arhitektura, 50; Leksikon 3, 290-291 .
, , 57.
For a general chronological frame of the development of Bosnia, see: .
, I, 1940; . ,
, 1964.
158
Dejan Buli
Croatia
Most of the present-day Croatia belonged to the province of

Dalmatia, with the exception of the northern, flat areas that were parts of
the Upper and Lower Pannonia, i.e. the provinces of Savia and Pannonia
II. Byzantine presence in Slavonia remains dubious. On the section of
limes from Aquincum to Singidunum, distance of several hundred kilometres, no Roman camp was discovered, not even in Mursa.331 The only relict
of urban life from the Late Antiquity is Siscia (Sisak), the town that survived until the early eighth century.332
Geographically speaking, the province of Dalmatia can be divided
into two areas, the coastal and the mountainous region. In the present time,
the coastal area belongs to Croatia, except for Neum. The littoral karst
region is characterized by a jagged coastline, shortage of drinking water,
and a few arable, fertile fields. There are only few passages fit for travel in
the high, insurmountable mountains immediately beyond the coastline.
Only two existed through the mountain Velebit the northern one,
through which Senj was connected with the Iapyd lands in the present-day
Lika and with Sisak; and the southern one, which connected Lika with
Ravni Kotari. Except for these, the passage from Klis to Sinjsko polje led in
the same direction as did the communication line along the Neretva river.333
Roman roads built in the early first century AD, immediately after
the conquest of these lands, facilitated the control and the process of
Romanization in Dalmatia and Illyricum. The proximity of the Adriatic
seaports made the delivery of material and goods, required by the army,
convenient. A string of permanent Roman camps was erected in the area
stretching from the Krka to the Neretva rivers, and south of the Dinara
mountain. Among these, only two legion camps stood: Burnum and
Tilurium, while auxiliary camps were based in Promona, Magnum,
331
332
333
M. Sanader, Rimske legije i njihovi logori u hrvatskom dijelu panonskog

limesa, Opuscula archaeologica 27 (2003) 463-468.
B. Miggoti, Arheoloka graa iz ranokranskog razdoblja u kontinentalnoj
Hrvatskoj, Od nepobjedivog sunca do sunca pravde. Rano kranstvo u
kontinentalnoj Hrvatskoj, Zagreb 1994, 47.
J. J. Wilkes, Dalmatia, London 1969, XXI-XXVII.
159
Andetrium and Bigeste.334 After the conquest of Dalmatia, the population

came down from strongholds (gradinas) into the plains and foothills,
where antique settlements developed. Antique settlements, which existed
until the fourth century were situated near a moderately hilly terrain, on
slightly lifted terraces in the middle of fertile plains, close to the sources of
fresh water and yet safe from seasonal floods.335
In the turbulent times of the Late Antiquity, these prehistoric
locations were revived and turned once again into fortified settlements.
The frequent barbarian incursions that move in from the north and used
the roman roads forced the endangered and decimated population to seek
protection in these fortified sites that then evolved into genuine settlements. This pattern of life became a habit out of necessity, not because
these sites served as shelters, which they did not. The process of the socalled horizontal migration took place in the coastal region of Dalmatia, in
which the inhabitants of the coastal area moved to the islands and maintained contacts with the mainland via the sea.336
Within the class of fortifications from the Late Antiquity, focus
in Croatia was only on the fortifications erected on promontories and
towering heights of certain islands, and in similar locations on the coast
line. Some of these structures were built on uninhabited islands, or in locations far from any settlements, which led to the conclusion that they were
not built for defensive purposes, but that they together formed a system
that was meant to ensure full control over seafaring on the eastern coast of
the Adriatic. Their position to each other and to the main seafaring routes
between the islands and along the coast point to this, too.337
Zlatko Gunjaa classified the Late Antique fortifications on the
coastline and on the islands. Besides the fortifications he assorted with
utter certainty, he also mentioned the positions in which remains of forti334
335
336
337
D. Peria, Je li delmatsko podruje presjekao rimski limes?, Archaeologia

Adriatica 2 (2008), 507; I. Borzi - I. Jadri, Novi prilozi arheolokoj
topografiji dugopoljskoga kraja, Archaeologia Adriatica 1 (2007) 167.
T. Tkalec, S. Karavani, B. iljeg, K. Jelini, Novootkrivena arheoloka
nalazita uz rjeicu Veliku kod mjesta Majur i Ladinec, Cris. asopis
Povjesnog drutva Krievci 9-1, Krievci 2007, 5-25.
. Tomii, Arheoloka svjedoanstva o ranobizantskom vojnom
graditeljstvu na sjeverojadranskim otocima, Prilozi 5/6 (1988/1989), Zagreb
1990, 29-53.
Z. Gunjaa, Kasnoantika fortifikacijska arhitektura na istonojadranskom
priobalju i otocima, Odbrambeni sistemi u praistoriji i antici na tlu
Jugoslavije, Materijali 22, Novi Sad 1986, 124 (= Gunjaa, Kasnoantika).
160
Dejan Buli
fications allegedly existed (but were yet to be confirmed), some positions

which he marked based on his own impressions, the importance of the
locations and the potential oversight over seafaring in a wider area.338 From
this list and from the fortifications provided by Goldstein,339 here were
included only those that underwent archaeological excavations as well as
those where architectural elements have been preserved. Count of the
already-mentioned fortifications from the Late Antiquity/Early Byzantine
period we added to the fortifications in the hinterland of Dalmatia, as well
as those covered by the latest excavations, to the extent of availability of
more recent publications:
1. Fortifications on the cape Molunat (15th century)340
2. pidaurus (Cavtat) (up to the 9th century, Late Middle Ages)341
3. Island of Mrkan342
4. Islet of Bobara near Cavtat343
5. Gradac near Dubrovnik344
6. Spilan above upa at Dubrovnik345
7. Dubrovnik (continuity)346
8. Stari Grad in the Peljeac peninsula347
9. Fortifications on St. Micheals hill in Peljeac (church, 11th century)348
338
339
340
341
342
343
344
345
346
347
348
Such assumptions are supported, in some cases, by the toponyms of these

sites, or by the continuous presence of fortifications on them, whose
construction most probably destroyed previous structures: Gunjaa,
Kasnoantika, 128-129.
Goldtajn, Bizant.
L. Bereti, Molunat. Utvrde i regulacioni plan Molunata iz druge polovine
15. stoljea, Prilozi povijesti umetnosti u Dalmaciji 14, Split 1962, 53;
Gunjaa, Kasnoantika, 128; Goldtajn, Bizant, 34.
Sui, Antiki grad, 35; Goldtajn, Bizant, 34.
I. Fiskovi, O ranokranskim spomenicima neronitskog podruja, Dolina
rijeke Neretve od prethistorije do ranog srednjeg vijeka, Izdanja HAD 5,
Split 1980, 243 (= Fiskovi, O ranokranskim); Gunjaa, Kasnoantika, 128;
Goldtajn, Bizant, 34.
Fiskovi, O ranokranskim, 249; Gunjaa, Kasnoantika, 128; Goldtajn,
Bizant, 34.
I. Marovi, Arheoloka istraivnja u okolici Dubrovnika, Anali Dubrovnik 45 (1955/1956) 9-31 (= Marovi, Arheoloka istraivanja); Goldtajn, Bizant, 34.
Marovi, Arheoloka istraivanja, 24; Goldtajn, Bizant, 34.
Gunjaa, Kasnoantika, 128; Goldtajn, Bizant, 36-37.
M. Zaninovi, Antika osmatranica kod Stona, Situla 14/15, Ljubljana
1974, 163-173; Gunjaa, Kasnoantika, 125; Goldtajn, Bizant, 38.
C. Fiskovi, Likovna batina Stona, Anali Dubrovnik 22-23 (1985) 80;
10. Polaa in ljet349

11. Katel in Mljet350
12. Fortification in the upper part of the islet of Majsan351
13. Fortification in the site Glabalovo selo above Orebi352
14. Straa above Pjevor in Lastovo353
15. Fortification on the islet of Svetac, near Vis354
16. Gradina above Trpanj in Peljeac355
17. Zamasline in Peljeac356
18. Baina at Ploe357
19. Fortification on the island of Osinje358
20. Gradina in Jelsa359
21. Faros-Starigrad (continuity)360
22. Grad or Galenik on the hill Paljevica, in Hvar361
23. Tor in Hvar362
24. Fort Graee on the exit out of Starigradski bay363
25. Bol on the island of Bra (9th century)364
26. Mirja above Postire in Bra365
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
M. Sui, Antiki grad na istonom Jadranu, Zagreb 1976, 239.

Gunjaa, Kasnoantika, 125
Gunjaa, Kasnoantika, 125.
Fiskovi, O ranokranskim, 230; Goldtajn, Bizant, 39.
B. Kirgin - A. Miloevi, Svetac, Arheo 2, Ljubljana 1981, 45-51; Gunjaa,
Kasnoantika, 125; Goldtajn, Bizant, 40.
I. Fiskovi, Peljeac u protopovijesti i antici, Peljeki zbornik 1, Zagreb
1976, 15-80; Gunjaa, Kasnoantika, 125; Goldtajn, Bizant, 42.
Fiskovi, O ranokranskim, 221; Goldtajn, Bizant, 42.
Fiskovi, O ranokranskim, 14-15; Goldtajn, Bizant, 42.
J. Jelii, Narteks u ranokranskoj arhitekturi na podruju istonog Jadrana,
Prilozi povijesti umjetnosti u Dalmaciji 23, Split 1983, 26-27; Gunjaa,
Goldtajn, Bizant, 42; M. Kati, Nova razmatranja o kasnoantikom gradu na
Jadranu, Opvscula archaeologica 27 (2003) 525 (= Kati, Nova razmatranja).
On the Croatian coast Faros is the only example of a town from the
Antiquity that underwent a reduction in its urban form: Kati, Nova
razmatranja, 525; Goldtajn, Bizant, 42-43.
M. Zaninovi, Neki prometni kontinuiteti u srednjoj Dalmaciji, Materijali
17, Pe 1978, 39-53; Goldtajn, Bizant, 43.
D. Hrankovi, Braciae insulae descriptio (Opis otoka Braa), Legende i kronike,
Split 1977, 210, 219; Goldtajn, Bizant, 43.
E. Marin, Mirje nad Postirama, AP 19 (1977) 152-154; Goldtajn, Bizant, 43.
161
162
Dejan Buli
27. Salona366
28. Split (Diocletians Palace) (continuity)367
29. Trogir (continuity)368
30. Gradina on the island of irje369
31. Gustijerna on the island of irje370
32. Tradanj on the lower Krka river371
33. St. Ana fortification in the ibenik area372
34. Fortification on the island of Vrgada373
35. reta arac on the island of Kornati374
36. Pustograd on the island of Paman375
37. St. Mihovil in Ugljan376
38. Koenjak near Sala in Dugi otok377
39. Graevina on the islet of St. Peter near Ilovik378
40. Jader (Zadar) (continuity)379
366
367
368
369
370
371
372
373
374
375
376
377
378
379

Goldtajn, Bizant, 44; T. Buri, Vinia. Rezultati rekognosciranja, SP 27
(2000) 59.
Gunjaa, Kasnoantika, 126; Z. Brusi, Kasnoantika utvrenja na otocima
Rabu i Krku, Arheoloka istraivanja na otocima Krku, Rabu, i Pagu i
Hrvatskom primorju, Izdanja HAD 13, Zagreb 1988, 111-119 (= Brusi,
Kasnoantika).
Z. Gunjaa, Gradina irje. Kasnoantika utvrda, AP 21 (1980) 133; Gunjaa,
Kasnoantika, 126; Brusi, Kasnoantika, 111-119.
Z. Gunjaa, O kontinuitetu naseljavanja na podruju ibenika i najue
okolice, ibenik. Spomen-zbornik o 900. obljetnici, ibenik 1976, 46 (=
Gunjaa, O kontinuitetu); Goldtajn, Bizant, 47.
Gunjaa, O kontinuitetu, 46; Gunjaa, Kasnoantika, 128; Goldtajn, Bizant, 47.
I. Petricioli, Toreta na otoku Kornatu, Adriatica Praehistorica et Antiqua
(ur. V. Mirosavljevi, et al.), Zagreb 1970, 717-725; Gunjaa, Kasnoantika,
126.
N. Jaki, Prilozi povjesnoj topografiji otoka Ugljana, Radovi FF-a u Zadru
15 (1989) 83-102; Goldtajn, Bizant, 48; Z. Kara, Tragovi bizantskog
urbanizma u Hrvatskoj, Prostor 3-2 (10), Zagreb 1995, 291 (= Kara,
Tragovi).
. Ivekovi, Dugi Otok i Kornat, Rad JAZU 235 (1928) 256; I. Petricoli,
Spomenici iz ranog srednjeg vijeka na Dugom Otoku, SP 3 (1954) 53-65;
A. Badurina, Bizantska utvrda na otoiu Palacol, Arheoloka istraivanja na
otocima Cresu i Loinju, Izdanja HAD 7, Zagreb 1982, 171-174; Gunjaa,
Kasnoantika, 128.
Goldtajn, Bizant, 49-50.
163
41. St. Damjan fortification in the island of Rab380

42. Katelin fortification above Kamporska draga on the island of Rab381
43. Fortification on the hill of Bosar, near Baka, on the island of Krk382
44. Fortification of Veli Grad on the cape Glavina, on Krk383
45. Fortification on the islet St. Mark (Almis)384
46. Gradina above Omilje, on the island of Krk385
47. Fortification on the islet of Palacol386
48. Apsorus (sor) (Late Middle Ages)387
49. Drid388
50. Island of Drvenik, at the foothill of Graina389
51. Ostrvica in Poljice390
52. Gradina above Modri draga391
53. Sveta Trojica392
54. Gradina above Donja Prizna393
380
381
382
383
384
385
386
387
388
389
390
391
392
393
Gunjaa, Kasnoantika, 127; Brusi, Kasnoantika, 111-119; Goldtajn,

Bizant, 51.
Gunjaa, Kasnoantika, 128; Brusi, Kasnoantika, 112; . Tomii, Sv. Juraj
iznad Paga. Ranobizantski kastron, Obavijesti HAD 21, Zagreb 1989, 28-31;
Kara, Tragovi, 293.
Gunjaa, Kasnoantika, 128; Faber, Osvrt, 116-121; Brusi, Kasnoantika,
112-116; Kara, Tragovi, 291; Goldtajn, Bizant, 52.
Gunjaa, Kasnoantika, 127; . Faber, Osvrt na neka utvrenja otoka Krka
od vremena prethistorije do antike i srednjeg vijeka, Prilozi 3-4 (1986/1987),
Zagreb 1988, 116-121 (= Faber, Osvrt); Brusi, Kasnoantika, 111-119;
Kara, Tragovi, 291.
N. Novak - A. Boi, Starokranski kompleks na Mirinama u uvali Sapan
kraj Omilja na otoku Krku, SP 21 (1991) 1995, 32.
A. Badurina, Bizantska utvrda na otoiu Palacol, Arheoloka istraivanja na
otocima Cresu i Loinju, Izdanja HAD 7, Zagreb 1982, 171-177; Gunjaa,
In the year of 530, it became the episcopal see: A. Faber, Poeci urbanizacije
na otocima sjevernog Jadrana, Arheoloka topografija Osora, Arheoloka istraivanja
na Cresu i Loinju, Izdanja HAD 7, Zagreb 1982, 61-78; Goldtajn, Bizant, 54.
M. Kati, Utvrda Drid, Prilozi povijesti umjetnosti u Dalmaciji 34 (1994), 5-19.
T. Buri, Arheoloka topografija otoka Drvenika i Ploe, SP 27 (2000), 41.
. Rapani, Kasnoantika palaa u Ostrvici kod Gata (Poljica), Cetinska
krajina od prethistorije do dolaska Turaka, Izdanja HAD 8, Split 1984, 149-162.
. Tomii, Matrijalni tragovi ranobizantskog vojnog graditeljstva u
velebitskom podgorju, Vesnik Arheolokog muzeja 23, Zagreb 1990, 139162 (= Tomii, Matrijalni tragovi).
A. Glavii, Arheoloki nalazi iz Senja i okolice (VI), Senjski zbornik 10-11,
Senj 1984, 17; Tomii, Materijalni tragovi, 139-162.
Tomii, Matrijalni tragovi, 139-162.
164
Dejan Buli
55. Kastron in Sutojanica (Svetojanj, Sutojanj, Svetojanica)394

56. St. Juraj above Pag395
57. Fortification on a plateau near Klopotnica396
58. Site Kolja Gromaa north of Novalja397
59. Trinielo near Stara Novalja398
60. Izvor near Kolan399
61. Fortification on the hill of Koljun near Zaglava (Novaljsko polje)400
62. Petri near Stara Novalja401
63. Fortification in Slatina above Gajac402
64. Gradina near Baka voda403
65. Site Luna in the western upper part of the island of Pag404
66. Guard post in the island of Ist405
67. Korintija on in the island of Krk (until the 11th century)406
68. St. Peter peninsula407
69. Beretinova gradina408
70. Hill Pupavica, in the Vuipolje area near Dugopolje409
71. Burnum, the Roman camp410
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
Tomii, Matrijalni tragovi, 139-162; . Tomii, Svetojanj. Kasnoantika

utvrda kraj Stare Novalje na otoku Pagu, Arheoloki radovi i rasprave 12,
Zagreb 1996, 291-305.
. Tomii, Arheoloka svjedoanstva o ranobizantskom vojnom
graditeljstvu na sjeverojadranskim otocima, Prilozi 5/6 (1988/1989), Zagreb
1990, 29-53. A Byzantine gold coin was discovered in one of the rooms: K.
Regan, Utvrda Sv. Jurja u Caskoj na otoku Pagu, Prilozi Instituta za
arheologiju u Zagrebu 19 (2002) 141-148 (= Regan, Utvrda). After the fall
under the Slavic control, the settlement kept on living until 1203, when it
was razed and deserted, during a conflict between Rab and Zadar.
Regan, Utvrda, 141.
Regan, Utvrda, 141.
Regan, Utvrda, 141.
Regan, Utvrda, 141.
Regan, Utvrda, 141.
Regan, Utvrda, 141.
Regan, Utvrda, 141.
Kati, Nova razmatranja, 523.
Regan, Utvrda, 141.
Kara, Tragovi, 291.
Kara, Tragovi, 290.
. Batovi, Istraivanje ilirskog naselja u Radovini, Diadora 4 (1968) 53-69.
I. Borzi - I. Jadri, Novi prilozi arheolokoj topografiji dugopoljskoga kraja,
Archaeologia Adriatica 1, Zagreb 2007, 160.
M. Zaninovi, Burnum, casellum-municipium, Diadora 4 (1968) 121; M.
Zaninovi, Od gradine do castruma na podruju Delmata, Odbrambeni sistemi
165
72. Knin, ancient Ninia411

73. Gradac (above the road leading to Promona), round the St. Marijen church412
74. Danilo Gornji, ancient Ridera near ibenik413
75. Balina glavica (Magnum)414
76. Gradina of Subotie415
77. Podgrae near Benkovac (Aserija) (Middle Ages)416
78. uker in Mokro Polje417
79. Keglevia gradina Mokro Polje418
80. Glavica near the small village of Meter in Lug (Middle Ages)419
81. Kokia glavica Pripolje420
82. Grad on the slopes above Knezovi and Mami jezero421
83. Ljubljan Ravni kotari422
84. Kuzelin near Zagreb423
85. Narona (Vid)424
411
412
413
414
415
416
417
418
419
420
421
422
423
424
u praistoriji i antici na tlu Jugoslavije, Materijali 22, Novi Sad 1986, 166 (=
Zaninovi, Od gradine).
M. Zaninovi, Kninsko podruje u antici, Arheoloki radovi i rasprave 7,
1974, 309; Zaninovi, Od gradine, 167.
A. Uglei, Ranohrianska arhitektura na podruju dananje ibenske
biskupije, Drni - Zadar 2006, 51-53.
M. Zaninovi, Gradina u Danilu i Tor nad Jelsom, Dva gradinska naselja u
srednjoj Dalmaciji, Materijali 15, Beograd 1978, 17-29 (= Zaninovi, Gradina).
I. Glava, Municipij Magnum. Raskrije rimskih cestovnih pravaca i
beneficijarska postaja, Radovi Zavoda za povijesne znanosti HAZU u Zadru
52, Zagreb - Zadar 2010, 45-59.
I. Alduk, Uvod u istraivanje srednjovekovne tvrave Zadvarje (1. dio - do
turskog osvajanja), Starohrvatska prosvjeta 32 (2005), 218.
Sui, Antiki grad, 136, including the relevant bibliography. Many structures
were dated to the Middle Ages.
Life on Gradina ended with the Slavic and Avar incursions, but several
ceramic fragments were discovered, dated to the Late Middle Ages: V.
Delonga, Prilog arheolokoj topografiji Mokrog Polja kod Knina, SP 14
(1984) 259-283 (= Delonga, Prilog).
Delonga, Prilog, 259-283.
Lj. Gudelj, Proloac Donji. Izvjee o istraivanjima lokaliteta kod crkve Sv.
Mihovila u Postranju, SP 27 (2000) 130. (= Gudelj, Proloac Donji)
Gudelj, Proloac Donji, 129-146.
Gudelj, Proloac Donji, 129-146.
Tomii, Materijalni tragovi, 147.
This fortification has existed since the 4th century: V. Sokol, Das spatantike Kastrum
auf dem Kuzelin bei Donja Glavica, Arheoloki vestnik 45 (1994) 199-209.
N. Cambi, Antika Narona. Postanak i razvitak grada prema najnovijim
arheolokim istraivanjima, Materijali 15, Beograd 1978, N. Cambi, Arhitektura
166
Dejan Buli
86. Gradina Badanj425

87. Bribir (Late Middle Ages, Ottoman period)426
88. Mala Vijola near Knin427
89. itluk near Sinj (ancient veteran colony of Aequum)428
This list enumerates 89 fortifications in Croatia, but this number

must have been higher. Until now, a plenty of strongholds (gradine) on the
territory of Mokro polje429 and dry-stone fortifications erected on the hills
overlooking Sinjsko polje have been sighted; some Late Antique/Early
Byzantine ones might be found among the latter.430 Just so, some fort
would surely be registered with sondages on a few of medieval fortifications on the slopes of Medvednica (Medvedgrad, Susedgrad), Samoborsko
gorje (Oki, Samobor), and umberak/Gorjanac (Mokrice).431 In the vicinity
of the already-mentioned Balina Glavica near Umljanovii (75), several
gradinas were discovered, some of which might be from the Early
Byzantine period.432
425
426
427
428
429
430
431
432
Narone i njezina teritorija u kasnoj antici, Radovi Filozofskog Fakulteta u

Zadru 24 (1984/1985) 33-58; E. Marin, Narona: Vid kod Metkovia, Split 1999.
Besides the Late Antiquity period, ranging from the fourth to the sixth
century, medieval findings were registered, dating from the thirteenth and
fourteenth centuries: R. Mateji, Gradina Badanj kod Crkvenice, Jadranski
zbornik 10, Pula 1978, 239-271.
Z. Gunjaa, Strateko i istorijsko-arheoloko znaenje Bribira, Kolokvij o
Bribiru. Pregled rezultata arheolokih istraivanja od 1959. do 1965. godine,
Zagreb 1968, 9-16; Z. Gunjaa, Nalaz srednjovekovnih arhitektura na
Bribiru, SP 10, Zagreb 1968, 235-242; T. Buri, Bribir u srednjem vijeku,
Split 1987.
M. Zaninovi, Kninsko podruje u Antici, Arheoloki radovi i rasprave 7,
Zagreb 1974, 303.
N. Gabri, Kolonia Claudia Aequum (Pregled dosadanjih iskopavanja,
sluajnih nalaza i usputnih zapaanja), Cetinjska krajina od prethistorije do
dolaska Turaka, Split 1984, 273-284. The town was mentioned in 533, at the
second Council of Salona; Sui, Antiki grad, 131.
Delonga, Prilog, 262
D. Peria, Je li delmatsko podruje presjekao rimski limes?, Archaeologia
Adriatica 2 (2008) 511-512; . Barlutovi, Neka pitanja iz povijesti Senja,
Senjski zbornik 34 (2007) 265-296.
D. Lonjak - Dizdar, Terenski pregled podruja izgradnje HE Podsused,
Annales Instituti Archaologici 4 (2008) 109-112.
I. Glava, Municipij Magnum. Raskrije rimskih cestovnih pravaca i
beneficijarska postaja, Radovi Zavoda za povijesne znanosti HAZU u Zadru
52, Zagreb-Zadar 2010, 45-59.
167
By applying the criteria of urban continuity, Z. Kara proposed the

following classification:
- Towns with antique foundations
- Dislocated, i.e. abandoned towns
- Newly-emerged settlements, some of which lasted continuously433
According to the proposed classification, Zadar (40), partly Trogir
(29) and probably Rab, too (41-42) fall into the first type of settlement, i.e.
they represent towns with the least turbulent transitions from the
Antiquity and Byzantine era to the Middle Ages.434 These towns survived
historic calamities, but have persevered up to the present, and are towns
with full continuity of existence.
The second group of settlements are those that transferred their
urban functions to more secure areas towards the coast or to the islands
when the hinterland was lost and the terrestrial communication interrupted.
The dwindling population of Salona (27) moved closer to the sea partly
into Diocletians palace (28), from which the town of Split would develop,
and partly towards the nearby Trogir.435 The population of Epidaurus (2)
sought refuge on the nearby islands of Mrkan (3) and Bobara (4),436 but also
to the gradinas of Gradac (5) and Spilan (6), that had already been inhabited for centuries before,437 while the episcopal see was transferred to
Dubrovnik (7). Epidaurus lingered on until the ninth century.438 Narona,
an important harbour on the Neretva, was transferred above Ston (8-9)
when the lower course of the river silted;439 the same phenomenon struck
Nin (Aenona) too.
But some ancient cities disappeared completely because new locations could not be found, which happened to a whole string of settlements
433
434
435
436
437
438
439
Kara, Tragovi, 285-298.

Kara, Tragovi, 285.
Goltajn, Bizant, 91.
I. Fiskovi, O ranokranskim spomenicima naronitskog podruja, Dolina
rijeke Neretve od prethistorije do ranog srednjeg vijeka, Izdanja HAD 5,
Split 1980, 233, 246, 249; Goldtajn, Bizant, 34.
Annales Anonymi Ragusini, Monumenta spectantia historiam Slavorum
meridionalium 25, Zagreb 1983, 7; Goltajn, Bizant, 34. For further information
regarding the results of the archaeological excavations, see: I. Marovi,
Arheoloka istraivanja u okolici Dubrovnika, Anali Dubrovnik 4-5 (1955-6)
24; J. Medini, O nekim kronolokim i sadrajnim znaajkama poglavlja O
Dalmaciji u djelu Cosmographia anonimnog pisca iz Ravene, Putevi i komunikacije u antici, Materijali 17, Pe 1978, 76-77 (= Medini, O nekim kronolokim).
Kara, Tragovi, 289.
Goltajn, Bizant, 96, 98.
168
Dejan Buli
below Velebit: Ortopla (Stitnica), Vegium (arlobag), Lopsica (urjevo),

Argyruntum (Starigrad). These settlements lost their terrestrial communications, and found themselves beyond Byzantine sea routes.440 Senia (Senj)
was the only town to have arranged transfer of its location to the castrum
of Korinthia on the island coast of Krk (67), which lasted until the
eleventh century.441
Late Antique underwent transformations, due to historical events
and economic factors, political and administrative changes, and new
cultural and ideological structures, as analysed in detail by M. Sui.442
Towns underwent ruralization; elements of rural economy and rustic
architecture spread inside towns elements of agrarian production in
urban palaces.443 Most of the agglomerations inherited from the Antiquity
were ruralized and thus survived in the form of the agro-urban
milieu.444 In the Late Antiquity, towns were depopulated and villages
repopulated. This exodus of the urban population was a consequence of
the permanent economic crisis, which led to the growth of villages and of
agricultural production.445 In the hinterland, the prevailing insecurity
caused the strongholds to be re-evaluated; and not just the settlements
that continued to exist throughout the entire Antiquity, but also those
that were abandoned. This fits in the already stated tendencies of the
castrization process.446
Because the terrestrial communications were lost, the only road
stretching along the coast was the maritime one. Therefore, it was necessary to build a system of watchtowers and fortifications along the sea
routes of Byzantine ships. Around forty of them were built in the area
stretching from the cape Planka in central Dalmatia to the coast of Istria,
5-10 km apart, allowing for visual communication.447 The other reason to
440
441
442
443
444
445
446
447
Kara, Tragovi, 289.

Kara, Tragovi, 289-290.
For further information regarding the transformation of antique towns into
medieval ones (post-Antiquity), changes and reduction of public spaces,
construction and adaptation works, usage of monumental objects for
secondary purposes, issues regarding spolia, spacial conceptions, internal
disposition and articulation, as well as the questions of spatial solutions and
relations within a town area, see: Sui, Antiki grad, 227-251.
Sui, Antiki grad, 248-9.
Sui, Antiki grad, 248-9.
Sui, Antiki grad, 249.
A. Badurina, Bizantski plovni put po vanjskom rubu sjevernih jadranskih
otoka, Radovi Instituta za povijest umetnosti 16, Zagreb 1992, 7-9.
169
construct fortifications was to create a neccessary network of refuges for

the adjacent unfortified rural settlements.448 Most of these fortifications did
not survive the Middle Ages, although some of them stood for a very long
time, like the Brioni castel, which existed until the sixteenth century. The
fortified Byzantine locations were abandoned early, especially the
agglomerations on high altitudes, far away from the sea and/or a suitable
harbour. Late Antique rural palaces (e.g. strvica in Poljaci, Polaa on
Mljet) suffered a similar faith. In the sixth and seventh centuries, small
rural settlements of a limited duration formed around them.449
Most authors attribute the horizons of fortification creation
along the eastern Adriatic coast to the reconquista of the Emperor
Justinian.450 These fortifications were, doubtlessly, providing safety for the
naval transportation in this part of the Adriatic, bays suitable for anchoring and safe from winds establishing control over the navigation routes,
and were offering protection to the local population. Byzantium showed
significant interest in harbours and islands lying on the east Adriatic coast,
since that route enabled the most direct and, in the aftermath of Slavic
migrations, the only connection with Ravenna and the territories in
northern Italy. Pursuing the goal of the restoration of the Roman Empire
within its former boundaries, Justinian had to defeat the Gothic fleet. And
only after the victory was won, at the beginning of the second half of the
sixth century, Byzantium managed to seize the entire Adriatic. According
to Gunjaa, this period should be considered terminus post quem for the
start of the construction of the fortification system, at least regarding the
structures in the central and northern parts of the Adriatic.451
Contrary to the aforementioned prevailing opinion of the utter
demise of Illyricum (depopulation, ravaged economy, razed and abandoned towns as a consequence of the plague epidemics, loss of trade and
traffic connections with the occupied hinterland), Kati considers the
process of decay and transformation of the Dalmatian towns to be far more
complex and lengthy in nature. Recent research has also pointed to another, quite opposite process that took place in the Late Antiquity.452
448
449
450
451
452
Goltajn, Bizant, 104.

Kara, Tragovi, 294.
Tomii, Matrijalni tragovi, 146; Kara, Tragovi, 291; Regan, Utvrda, 147.
Gunjaa, Kasnoantika, 131.
M. Kati, Nova razmatranja o kasnoantikom gradu na Jadranu, Opuscula
archaeologia 27, Zagreb 2003, 523-528 (= Kati, Nova razmatranja).
170
Dejan Buli
The process of decay indeed struck larger towns, like Salona and
Narona.453 The author underlines the example of Hvar, i.e. Lisine, founded
at the end of the fourth century. It had no earlier roots in the Antiquity,
and yet it flourished in the sixth century.454 Kopar and Novigrad in Istria
have roots in the Late Antiquity, and the same applies to Biograd, ibenik
and Dubrovnik.455 Written sources and archaeological excavations clearly
indicate that new fortified centres of the Late Antiquity contain ports,
churches and ramparts, and some of these became diocese sees. Because of
all this, Kati claims that the notion of the crisis of the Late Antiquity in the
eastern Adriatic needs to be more clearly defined, depending on the available archaeological and historical facts, which is why generalizing the process
of urban settlements decay cannot be accepted.456
Cosmographia of the Anonymous from Ravenna speaks in favour
of this hypothesis. In this work, the number of towns registered compared
to the earlier Roman itineraries is higher. These are the new centres of the
Late Antiquity,457 and the newly-established system of habitation in the littoral
regions.458 The anonymous writer from Ravenna, author of Cosmographia,
a treatise composed at the end of the sixth or in the early seventh century,
designated civitates on the coastal stretch of land at the foothill of the
Velebit Mountain.459 The explanation given for this fact was that his contemporaries did not differentiate between towns and villages (and even
today many rural settlements are called towns), and that the fortifications
held so much importance that a mere presence of ramparts enhances the status of settlement.460
As Slobodan ae states, the accounts given by the Anonymous of
Ravenna are precious as they indicate that the process of castrization - i.e.
transfer of settlements towards more easily defensible hilltops had advanced
well even before the sixth century and that it took off during Justinians
453
454
455
456
457
458
459
460
Goltajn, Bizant, 90-91; 96.

M. Kati, Kasnoantiki grad na Jadranu. Primer grada Hvara, Prilozi povijesti
umetnosti u Dalmaciji 38, Split 1999/2000, 19-49 (= Kati, Kasnoantiki grad).
Medini, O nekim kronolokim, 69-83.
S. ae, Civitates Dalmatiae u Kozmografiji Anonima Ravenjanina, Diadora
15 (1993) 431 (= ae, Civitates Dalmatiae).
Ravennatis Anonymi Cosmographia et Gvidonys Geographica (ed. M.
Pinder - G. Parthey), Berlin 1860; Sui, Antiki grad, 303-305.
More extensively on this issue, see: Sui, Antiki grad, 248-9; Goldtajn,
Bizant na Jadranu, 101-2.
171
reign, when the entire province was covered by a dense network of

different fortifications, ranging from towns and small fortifications to
watchtowers. Even the settlements that were not on the main road were
listed: Dubrovnik, Ston, Makar, Drid; 461 but still, many settlements were
located along the sea routes, which should be taken into consideration.
With the cessation of terrestrial traffic, commerce and traffic
shifted to sea routes. Skilful in seafaring and shipbuilding, the islanders
benefited from the newly-emerged circumstances and took part in trade
and transportation in the Mediterranean. These circumstances led,
together with an increasing influx of population, to the formation of late
antique civitates on the coastline. In this process, Justinians reconquista
played an important, but not the key role.462 Justinian was not setting up
a limes by building fortifications along the sea roads of the eastern
Adriatic, but was rather striving to secure and improve the seafaring
conditions in the Adriatic.463 Therefore, castrization was not the only
process taking place, but also the construction of docks and harbours,
usually below fortifications, which was helping trade and providing
harbours for ships.464
Tomii and several other authors had an idea of a limes set up
along the southern coast below the Velebit mountain, acting as a defensive
system against Slavic incursions towards the Adriatic;465 I. remonik
shared this opinion to a certain extent.466 Such a point of view is a product
461
462
463
464
465
466
ae, Civitates Dalmatiae, 430.

Kati claims that the impact of Justinians castrization, with which the Late
Antiquity fortifications on the eastern Adriatic coast are associated, has
been overstimated, and that earlier, fortified settlements on high terrain
need to be differentiated from the Early Byzantine castra: Kati, Nova
razmatranja, 525-526.
. Rapani, Predromaniko doba u Dalmaciji, Split 1987, 58.
The spatial distribution of the Early Byzantine fortifications lying at the
foothill of Velebit and on the island of Pag indicates their in-depth
arrangement, and that the forts on the coastal rim could have acted as the
first defensive line, with the castra on the Pag island being the second. The
position of mountain passages on the Velebit mountain fits with the
arrangement of the forts on the coastal rim, i.e. they are guarding the access
to the passages from the coast. The forts arranged in-depth on the Pags
coastal rim (56; 65) were guarding the naval zone, but also the island and its
urban agglomerations: Tomii, Matrijalni tragovi, 139-162.
She speaks of the clustering of fortified sites along the Adriatic coastal rim
and along the mountain ridges separating the coast from the Dalmatian
hinterland: remonik, Rimska utvrenja, 357.
172
Dejan Buli
of focusing on a small area only and of not perceiving the entire distribution of the fortifications, densely clustered throughout the territory of
Dalmatia and Illyricum.
The Avar forays and the Slavic colonization in the eastern
Adriatic, followed by the second wave of the arriving Serbs and Croats,
marked the end of Late Antiquity in these lands. In such circumstances,
the local Romanized population managed to survive for a long time in heterogenous enclaves surrounded by Slavs. It was only in Istria that nearly
all earlier settlements continued to exist,467 in contrast to very few on the
coastline of the present-day Dalmatia: Zadar, Trogir, Split and Dubrovnik
on the mainland, and Krk, Cres and Rab in the islands.468 The hinterland
was cut off, while the islands and the few surviving coastal towns maintained economic relations with the metropolis by the sea route. hese
were towns with an inherited continuity, cities with cultural, ethnic and
topical continuity (Krk, Osor, Rab, Zadar, Trogir..). Others preserved the
urban traditions of some destroyed town, but not its location, like Split and
Dubrovnik meaning, only cultural and ethnic continuity. More numerous are the settlements that rose at the sites of earlier urban settlements
from the Antiquity that suffered destruction, like Nin, Skradin and many
others. Although uninterrupted continuity has not been established in the
previously-mentioned cases, some precedents from the Antiquity played a
certain role in the formation of the new town, e.g. by retracing the ancient
ramparts, preserving some important communications...469 Some rural settlements would spurn urban organisation, even though they sprung up above
the antique ruins (Solin). Small Roman enclaves pressed against the coastal
rim could only have been rejuvenated by receiving fresh forces from the
hinterland. Thus began the process of Slavization in the coastal towns.470 As
we have seen in the afore-mentioned list, and as J. Medini said earlier, after
the Slavic colonization there were far more surviving Roman oases in littoral
Dalmatia than previously thought.471
While the issues regarding Byzantine towns on the eastern
Adriatic were widely spoken of and are now well-known, the Byzantine
fortifications in the hinterland remain a neglected topic. Because of this
we have today a very small number of fortifications in continental Croatia
467
468
469
470
471

DAI I 29. 49-54; , 12-13.
Medini, O nekim kronolokim, 75.
173
from this era, and the movable findings from such sites equally remain
unknown. It was already mentioned that the settlements in the rural areas
are developed along Roman roads, which now acted as the main streets.
Their locations in the valleys and the dispersion of dwellings made defensive features inexistent and fortification rather impossible, which is why
the population moved to the nearby hills and plateaus in tumultuous times
most often to the sites of former Illyrian strongholds.472 Except for a few
forts, they remain unfamiliar to us. If the analogous situation from the
nearest neighbourhood, Bosnia and Slovenia,473 is applied, an approximate
dispersion of Early Byzantine fortifications should be expected. Katis
remark on the re-use of Illyrian strongholds (gradina) provides a good
guideline for identifying them. A repeated analysis of the ceramic material could yield surprising results, by simply using the presence of mortar to
distinguish these epochs.
Montenegro
What is today Montenegro was, for the greatest part, the province
of Prevalis, which was detached from the province of Dalmatia at the end
of the third or beginning of the fourth century, in 297 or 305/6.474 All that
was said of the coastal towns in Croatia stands for those in Montenegro as
well. The towns in the maritime Zeta had a common origin - their urban
identity had been established in the Antiquity. Only a few of them
continued to exist into the Middle Ages without suffering destruction
during the Great Migrations (Ulcinj, Sva), while in the case of Acruvium,
still not localized precisely, the old settlement was entirely abandoned, and
a new one was founded that then took over the traditions of the old town
together with its diocese (Kotor). The town of Bar represents an exception,
since it was, most likely, restored in the sixth century, during Justinians
reconstruction of towns in Illyricum. All the coastal towns entered the
472
473
474

For further information regarding fortifications in Slovenia, see: S.
Cigleneki, Hohenbefestigungen als Siedlungsgrundeinheit der Spatantike
in Slowenien, Arheoloki vestnik, 45 (1994) 239-266; S. Cigleneki,
Hohenbefestigungen aus der zein vom 3. bis 6. Jh. Im Ostalpenraum,
Ljubljana 1987.
1, 1967, 242 (. ).
174
Dejan Buli
Middle Ages with a Christian population of Roman descent and as diocesan centres.475
While it is undeniable that the founders of the medieval towns in
the coastal Zeta were Romaions (Romanoi), the process of Slavization
began after the hinterland politically stabilized. In this process, the populations of the towns became mainly Slavic and the Romaions dissappeared
over time. Although greater or smaller Romaion islands persisted in the
towns, the urban districts were entirely Slavic.476 This process was followed by antagonisms between the native, Romaion population, and the
Slavic newcomers, which gained a sectarian note, in addition to the ethnic
one.477 Besides this, the Slavs in the hinterland lived of agriculture and animal husbandry, while the Romaions were forced to live of the sea.478
At the very beginning of the Early Medieval period, the episcopal
towns brought together the need for gathering, commerce, defence and
preservation of the Christian way of life. The last-mentioned is well
reflected in the fact that towns smaller in size and closer to each other
opened their doors for the refugees from the hinterland who carried with
them their dioceses (the cathedrae from destroyed Doclea and Acruvium
were transferred to Bar and Kotor, respectively).479
1. Bar (continuity)480
2. Ulcinj (Olcinium) (continuity)481
3. Old Ulcinj (Late Middle Ages, 17th century)482
4. Sva (8th-10th; 11th-15thcentury)483
475
476
477
478
479
480
481
482
483
. ,
IV XV , 2003, 17 (= , ).
, , 18-19.
. , --
, 14/1-2 (1958) 230.
, , 26.
, , 307.
. , , 1962; , , 42-44; .
, , 2008.
. - . - . , I, 1981;
, , 45-48; , , 124.
. - . , , 1975, 61, 141 (= - , ).
E. Zeevi, Late Phase of the Medieval Town Sva,
, 2001, 685-695; ,
, 27-33; 159; , , 48-50; ,
- (. .
), 2010, 249-250 (. ) (= ).
5. Budva (Buthua) (continuity)484

6. Duklja (Doclea) (church, 9thcentury)485
7. Gradac Budimlja486
8. Gradac Kaludra, Berane487
9. Gradina Andrijevica488
10. Onogot (Niki) (14th-15thcentury, Ottoman rule)489
11. Samograd (Kamengrad), in the vicinity of Berane490
12. Gradina uteza in Dinoe491
13. Vladimir (Oblak) near Sva492
484
485
486
487
488
489
490
491
492
-, , 162 ; , , 37-42; ,
, 34-36; 89-96.
Archaeologists M. ivanevi and D. Drakovi have confirmed the
existence of the Early Byzantine ceramics; P. Sticotti, Die rmischen Stadt
Doclea in Montenegro, Wiena 1913; P. Sticotti, Rimski grad Doklea u Crnoj
Gori, Podgorica 1999; - , , 63-69; Nova antika
Duklja I, Podgorica 2010; 1, 1967, 269-270
(. ); , , 160; Ceramical findings dated to the
earliest stage could be attributed to the 4th and 5th century, whereas the
traces of the 6th century are still questionable: D. Drakovi - M. ivanovi,
Keramika prostorije 3/IX. Prilog poznavanju svakodnevnog ivota antike
Duklje, Nova antika Duklja II, Podgorica 2011, 76-77.
. , ,
53-1, 1973, 117; . ,
,
(
17. 2010), - 2010, 182.
Based on an insight into the unpublished research of P. Lutovac,
archaeologist from the Polimski Museum in Berane.
1, 1967, 241-280, 253-4 (. );
- , , 122-123; , ,
187-188 (. ).
. , . ,
, 6 (1990) 135-139; .
, . , . A , 13
(1989) 31-46.
Given the provided description and the construction technique of the gradina,
we decided to include this site into the review, although excavations have
not been conducted. Cf. O. Velimirovi-ii, Ostaci fortifikacione arhitekture
na gradini uteza u Dinoama kod Titograda, Odbrambeni sistemi u praistoriji
i antici na tlu Jugoslavije, Materijali 22, Novi Sad 1986, 80-87, 14 (=
Velimirovi-ii, Ostaci).
, , 159-160.
175
176
Dejan Buli
14. Risan (Rhizinium) (Late Middle Ages)493

15. Herceg Novi (continuity)494
16. Nehaj (14th-16thcentury)495
17. Gradina Martinii (9th-12thcentury)496
As we have seen, like with the towns in Croatia, we can speak of
the continuity between medieval and antique towns in the case of the
coastal towns of Zeta. Residential and other buildings have not been
preserved, because later buildings were built of their material and on
their foundations. The earliest remnants of stone buildings belong to
sacral objects, inscriptions and stone carvings that allow the buildings to
be dated.497
Except for the typical towns, fortified places were also registered
(albeit to a lesser degree), such as fortified villages and occasional military
outposts that mostly did not survive into the Middle Ages. The threenaved basilica above Samograd from the Early Byzantine period is an
exception. At this site, fragments of medieval pottery dated to the tenth
century were discovered in the course of sondage exploration.498
Old medieval fortifications have been preserved in late medieval
towns thanks to their growth and development (Bar), or to their
stagnation (Sva). Fortifications from the eight to tenth centuries of other
towns have remained unknown, since they were either completely
demolished and built over or superposed on in the following period.499
493
494
495
496
497
498
499
- , , 130-131; , , 157-158;
, , 242-244 (. ).
A tower of a circular groundplan, now submerged under the sea but built
after the fall of the Roman Empire, indicates that a ferry traffic existed
between Lutica and Herceg-Novi: -, , 55-58;
P. Mijovi, Nekoliko opaanja o rekonstrukciji antikih i kasnoantikih
puteva kroz Crnu Goru, Putevi i komunikacije u antici, Materijali 17, Pe
1978, 133-144.
. , , 2008, 67-70.
- , , 61; V. Kora, Martinii. Ostaci srednjovekovnog grada, Beograd 2001; , , 51-55, 126-128.
For further information on this issue, including the bibliography, see:
, , 154-155.
rchaeological material has not been published yet: . , .
,
, 6 (1990) 135-139; . , . ,
. A , 13 (1989) 31-46.
, , 169.
177
According to the chapter XXXV of De administando imperio, three

inhabited towns existed in Doclea: Gradac, Novograd and Lontodokla, all
three still not located. Gradac could have been any of the many toponyms
with this name, but its name indicates its antique core, unlike Novograd.
The name Lontodokla is made out of two parts the latter being Dokla, i.e.
Duklja.500 Dioclea was too big to be a town in the Middle Ages, and no
reconstruction is known, which is why Lontodokla should be looked for
in the surrounding area.
Porphyrogenitos recorded the following inhabited towns in
Travunia and Konavle: Trebinje, Vrm, Risan, Lukaveti and Zetlivi.501 Since
Travunia extended into the areas of the present-day Montenegro up to Risan,
some of the mentioned towns were in the territory of the present-day
Montenegro, or they are assumed to have been. Risan is a well-known but
unexplored maritime town, lying on a hill approximately 200 m high. On
the other side of a stream, an older church was discovered beneath the
floor of the Sts. Peter and Paul church.502 Lukaveti and Zetlivi are unknown
and there are several proposals where to ubicate them.503
Jankovi identified Oblak from the Chronicle of Dioclea (Letopis popa
Dukljanina), with the site called Vladimir (Oblak), near Sva, where remains
of a church and of a fortification (dimensions 50 15/20m) were discovered.
The fortification was not inhabited throughout its whole existence, nor was
the refuge, as their surface areas are too small, so it must have served primarily
as a border-line fort. Fragments of antique pottery are the only movable
findings that were discovered. This site has not been researched.504
Life returned to the gradina of Dinoe at the time of Byzantine
restoration. Velimirovi-ii holds that this gradina could have been the
centre of Gorska upa.505 It is considered that these remains could represent
Novigrad from Porphyrogenitos writings.506
500
501
502
503
504
505
506
P. Skok, Kako bizantski pisci piu slovenska mjesna i lina imena, SP n.s. 1 (1927) 73.
DAI I, 34.19-20.
, , 158.
Cf. note 142. We mention them since these locations, most probably, had an
earlier, Late Antique phase.
, , 159-160.
Velimirovi - ii, Ostaci, 82-83. Two smaller gradinas guarded the rear of
the fortification; since these gradinas - in the north Gradac in Lopari, and in
the east, gradina Vuko - have not been researched, their chronological
frame cannot be established either.
The author identifies the remains of the medieval citadel as Ribnica. In support
of this hypothesis, he mentions the remains of the church of St. Archangel
178
Dejan Buli
In a string of castra appearing at the end of Antiquity, the fortified

settlement of Onogot should be underscored, built close to the former
Roman castrum Anderba in the present-day Niki in Montenegro. It is
believed it was named after a Gothic comes (Anagast, Anegast, hence
Anegastum), who had his residence at the spot.507
The small number of Early Byzantine fortifications comes as a consequence of never-conducted systematic reconnaissance on the one hand
and on the other, of a small number of excavations undertaken in the late
medieval towns that overlay the earlier strata. That it is so can be seen
from a simple fact that a large number of fortifications were discovered
around Berane after sondage works had been initiated.508 In this case, 11
out of 17 sites had later phases of use (64.7 %).
Macedonia
In 295, the reforms of Diocletian had Macedonia assigned into the

diocese of Moesia. In the first half of the fourth century, during
Constantines reign, it was transferred under the jurisdiction of the prefecture of Illyricum.509 he territory of this prefecture was divided into two
dioceses: Dacia in the north and Macedonia in the south. Macedonia was
made up of the following provinces: Macedonia Prima, Macedonia Secunda,
Epirus Nova, Epirus Vetus, Thessalia, Achaia and Creta.510 Macedonia I,
Macedonia II, and parts of the provinces Dardania, Dacia Mediterranea,
Praevalitana and Epirus Nova were situated within the boundaries of the
present-day Macedonia.511 The crisis that befell the Roman state and the
barbarian invasions affected Macedonia as one of Romes provinces.
The first Gothic incursions and the ravaging of towns in the third
century had a major impact on the eastern- and central-Balkan lands. The
507
508
509
510
511
Michael, in which Nemanja could have been baptized by the Catholic rite:
Velimirovi - ii, Ostaci, 82-83.
1, 1967, 253-4 (. ).
1, 1969, 53-54.
. , VI , 19 (1980), 19 (=
, ).
I. Mikuli, Sptantike und frhbyzantinische Befestigungen in
Nordmakedonien. Stdte-Vici-Refugien-Kastelle, Mnchen 2002, 19 (=
Mikuli, Sptantike).
179
area of Macedonia suffered destruction chiefly in 268 and 269, when most
towns, big and small, were destroyed and never rebuilt again. The barbarian menace reappeared after the ruin of the Roman army in the Battle of
Hadrianopolis in 387, when the Gothic squads ravaged the interior of the
Peninsula unchecked. In the fifth century, the barbarian threat became
the prime problem of the Balkan Peninsula. The Byzantine border on the
Danube repeatedly gave way in the mid-fourth century to the Hunnic
onslaught and the Ostrogothic incursions around 480. At the time, Stobi
and Heraclea Lyncestis were destroyed, Dyrrachium and Salona taken and
the surroundings of Thessalonica pillaged. Several barbarian assaults led by
the Avars, Bulgarians, Kutrigurs and Slavs struck the Balkans during the
sixth century. In 517, a barbarian squad made up of Geths (Bulgarians)
roamed Illyricum ultimately reaching Thermopylae, after plundering several forts on the way, Skupi among others.512 In 540, Huns (the Kutrigurs)
crossed the Danube and descended southwards to Chalkidiki. During this
raid, 32 fortified sites in Illyricum were destroyed.513
Taking a lesson from the experiences with the Huns and the
Goths, Emperors Leo and Zeno, followed by Anastasius and Justinian,
conducted fortification efforts to restore Late Antique fortifications and
to construct many new ones. Procopius of Caesarea compiled a list of
fortifications that were restored and towns that were built in provinces
and smaller regions; he made a record of 47 newly-erected and restored
forts in Macedonia.514 ustinians defensive system did not withstand the
Avaro-Slavic incursions in the years that followed. During the Kutrigur
raid of 558/9 that the Danubian Bulgars and Slavs joined, Lower Moesia
and Thrace were devastated and one of their parties proceeded towards
Thessalonica. Several fortifications in Macedonia were most likely
destroyed in this raid. Twenty years of peace followed, except in 571,
when the Slavs (judging by a horizon of deposits) penetrated all the way
to Macedonia.515 During the 580s, Slavic invasion from the lower Danube
512
513
514
515
. , ,
1996, 24 (= , ).
Proc. BG II 4, 163.8-164.16.
1, 59.
V. Popovi, Une invasion slave sous Justin II inconnue des sources crites,
4, 1981, 111-126. In Voden by Skoplje, on the
acropolis tower, a stratum with traces of fire and demolition was
stratigraphically established: . ,
, 1982, 50-51.
180
Dejan Buli
overran Thrace, parts of Illyricum (mainly those in Macedonia), and then

spread southward, to Hellas and Peloponnese. This campaign grew to
become a permanent barbarian settlement. Cases of hoarding soared,
indicating the jeopardy or disappearance of earlier urban life in the towns
of the mid-580s Macedonia.516 The next Slavic incursion happened
probably in 580/81, when the Slavs penetrated deep to the south, into
Greece, where they spent the following four years (581-584).517 Some areas
of Macedonia undoubtedly suffered destruction at that time, too. Already
in 584/5, the Avars from the Valachian plain, combined with the Slavs
from the Ukraine and Moldova, thoroughly desolated eastern Balkan
provinces and reached Thessalonica in 586.
The remaining Romaion population fled either to the south or into
the inaccessible mountain fortifications. The Byzantine presence was
reduced to holding a few most important fortifications. Evidence of
continuous life was discovered in about twenty larger fortified sites: coins
from the end of the sixth and the beginning of the seventh century, and a
Byzantine soldier fibula from the same period.518 After the collapse of the
limes lines in erdap, the Slavs gradually occupied entire Greece in waves
of settlement, having already covered Macedonia on their way.
In the territory of the present-day Former Yugoslav Republic of
Macedonia, more than 500 fortifications were registered. They were all
dated to the Late Roman (mid-200s mid-400s) and the Early Byzantine
period (late fifth sixth centuries).519 Their number is significantly higher
than the available written sources tell us.520 The information on most of the
fortifications comes from summary reports or side notes, since only very
few of them underwent thorough archaeological research. Eighty one
medieval fortifications with an antique nucleus have been enlisted. Of
516
517
518
519
520
, , 20-48.
V. Popovi, Aux origines de la slavisation des Balkans. La constitution des
premirs Sklavinies macdoniennes vers la fin du VIme sicle, Comptes
rendus de lAcadmie des inscriptions et belles-lettres I, Paris 1980, 232.
More extensively on the settlement of the Slavs into the Peloponnese and
Greece, see: . , ,
2002, 65-83; 119-141 (= , ).
, , 26.
For further information regarding the proposed categorization of the
fortified sites, see: . , , ,
1999, 190-191(= , ).
Procopius list omits in entirety the provinces of Prevalitana and Macedonia
II, and a greater part of Macedonia I.
181
these 81 registered medieval fortifications, the existence of the previous

stage (Late Antiquity) was found lacking only in the case of one town in
Macedonia, Debar:
1. Gradite Budinarci (Budingrad?)
2. Bitola Herakleja Linkestidska (Heraclea Lyncestis)
3. D. Oreovo Kale
4. ivojno Gradite
5. Zovik (emren) Gradite
6. Streevo Kale (Gabalarion?)
7. Belica
8. Belica, site Kale na Stolovatec
9. Devi, site Devini Kuli
10. Zagrad (Raste)
11. Zdunje
12. Iite, Vulkanska kupa Kale
13. Modrite, site Markovo kale
14. Valandovo
15. Vinica, site Gradite, Kale
16. Gabrovo (Petrovo), site Markov manastir
17. G. Banica, site Gradite
18. Gradec, site Gradite
19. Kalite, Grad Sokolec, site Sokolec
20. Srbinovo (Trnovo), site Kale Zvezda
21. Ratani Kale
22. Vire
23. Drame (Bigla)
24. Zvegor, site Malo (dolno) Gradite
25. Lukovica
26. Buin Kale
27. Graite Gradite, medieval Dobrun
28. eleznec Gradite
29. Drenovo, site Gradite, or Devol-grad
30. Resava
31. rangel (Srbica)
32. Kievo
33. Podvis
34. Morodvis, site Gradite
182
Dejan Buli
35. Opila, site Gradite

36. Gradite, the town of Lukovo?
37. Kanarevo, the town of Kozjak, site Grade
38. Konjuh, site Golemo Gradite
39. Mlado Hagoriane, the town of egligovo
40. Demir kapija, Prosek
41. Markov grad Korenica
42. Kula Korenica
43. elovec, site Strezov Grad (kale), the village of elovec
44. Godivje Kula, site Kula
45. hrid
46. Pesoani Kula, Debrica (Deuritsa)
47. Varo, the town of Prilep, site Markovi Kuli
48. Debrete, site Kale
49. Desovo, site Kale, Leska
50. Zrze, site Kale, Sveti Spas
51. Manastir Gradok, Markovi Kuli (the town of Morihovo?)
52. Prilepac, site Markov zid
53. Treskavec
54. Zletovo, site Bauar, Gradite
55. Radovi, site Hisar
56. opur, Brdo Pilat Tepe
57. Evla Kale, Vasilida?
58. Stenje (Konsko) Golem Grad (Golema Petra)
59. Trebenite Kale
60. Sopot (Trstenik), site Donjo Gradite
61. Vodno, the town of rne, site Markovi Kuli
62. Kole, site Markovi Kuli
63. Markova Suica, site Markovi Kuli
64. Matka, site Markov grad
65. Skopje, site Kale
66. uer, site Davina or Kula
67. Zagradani
68. Bansko, Termica? site Gradite
69. Kone, site Gornja Kula
70. Kosturino (Raborci), site Vasilica
71. Strumica, Ridot Careva Kuli
72. Donja Lenica, site Kale
183
73. Jegunovce, Gradite

74. Leok, site Kale or Gradite
75. Orae, site Gradite, Sobri
76. Rogle, site Kuka
77. Stene, the town of Stena?, site Gradite
78. Teovo, site Markovo kale
79. Veles, site Kale
80. Krupite, site Kale
81. Creka, site Hisar
82. tip, site Hisar521
Expressed in percentages, 16.4% of antique forts had a medieval
town or a medieval fortified site appearing after them. A conclusion can be
made that new stone fortifications were seldom encountered in medieval
Macedonia and are therefore an exception.
I. Mikuli proposed a categorization of medieval fortified sites
that we convey here without questioning its accuracy: castrum, regional
centre, mining fortification and the settlement, guard, refuge, fortified
monastery and suburb-town.522
Ramparts on some of the fortifications were considered in a good
shape. This was of particular importance in the Middle Ages, when it was
necessary to repair only the dilapidated upper parts of the ramparts, battlements, towers, gates and so on. Because these sections caved in at some
later time, it is difficult to register construction interventions everywhere.
This is why the remark of I. Mikuli that some fortifications were not
rebuilt in the Middle Ages, does not hold up, since these reconstruction
works could not have been registered. Examples of walling-off are easier
to notice but are less frequent, too. Most often, the walling-off was part of
resizing and reducing the fortifications, and the settlements were made of
timber and are therefore undetectable except by excavations. Having this
in mind, rare medieval findings make sufficient evidence that the fortifications were used in the Middle Ages.
With the Slavic settlement, and then with the arrival of the Serbs
and the Croats, most of Illyricum became colonized by Slavic tribes and
521
522
The list of sites was composed after the following work: ,

, with the exception of the site 21, which was
taken from the work: Mikuli, Sptantike, 278; and the site 59, taken from:
, , 405.
, , 134.
184
Dejan Buli
removed from the Byzantine control. However, there are no historical

accounts on these first decades and relations between Byzantium and the
sclavinia in Macedonia.
There were several attempts to deal with the situation in the
Western areas of the Empire but this resulted only in a temporary
subjugation to the supreme authority. Despite the transfer of populations
to Asia Minor, nothing resulted in a permanent solution, as the Slavic
sclavinia were rapidly acquiring independence. The restoration of power
in Thessaly and Macedonia began during the reign of the dowager-empress
Irine, when the Byzantine army led by the logothete Stauracius defeated
and subjugated the Slavs in 783. The introduction of the theme system
began after this event.523 The theme of Macedonia was mentioned already
in 802,524 and the theme of Thessalonica was created in the years
immediately following the campaign of Stauracius.525 What followed was a
massive Christian Romaion colonization in the areas of Strimon, especially
in 810/11, during the reign of Nicephoros. The aim was to Hellenize the
Slavs and reduce the threat from the ascending Bulgaria under Krum.
It should be logical to expect that the sclavinia were not taken
without a fight and that therefore some Early Byzantine fortifications
were then used for defensive purposes during the Byzantine offensive in
the Slavic territories.
No matter how much of a target or a stopover Macedonia was for
the foraying Slavs, traces of their presence are scarce in the seventh, eighth
and most of the ninth century. But we cannot agree with the opinion that
the Slavs came from the valley of Danube simply passed through Macedonia,
(already ravaged, with no traces of Romaion settlements, showing how
unattractive it became) and settled in the coastal Mediterranean Greece
with a mild climate. This does not seem credible.526 The accounts on the
founded sclavinia, together with the Early Slavic findings from the basin
of the Bregalnica river refute that theory. A ceramic vessel dating from the
seventh century was discovered on the left bank of the Bregalnica, while
fragments of an urn dating from the seventh century were unearthed on
the site Kazandijska Mala in tip. Ceramic fragments of hand-made pottery
discovered at Berovo and dating from the seventh and eighth centuries
523
524
525
526
For further information regarding the issue of sclavinia in Macedonia, see:

, , 204-239.
Teoph. I, 475.22.
, , 232.
, , 27.
185
indicate the presence of a Slavic ethnic group around the upper course of
the river Bregalnica. A bronze casting mould of Avaro-Slavic type was
discovered at the site Bargala by tip in a role of an amulet and as part of
funerary inventory. It was dated to the early seventh century.527
However, the few findings from the fortified sites, dated to the
seventh and eighth centuries indicate the presence of a non-Slavic
population. These were attributed to the autochthonous Romaions, bearers
of an inherently non-Slavic culture, as the case of the Komani-culture
necropolis beside Ohrid confirms. On the island of Golem-grad in the
Prespa lake, tombs were discovered containing jewelry of Byzantine-Italian
type and coins of Constantine IV (668-685); whereas coins of Constans II
were discovered at Isariot near Valandovo and at Selce near Prilep, as well
as the coins of Justinian II (685-695) at the acropolis of Konjuh. In the castle
of Debrete near Prilep, objects of Byzantine origin were discovered and
dated to the seventh century.528 Although these findings are not a priori
proof of Romaion in the most important fortifications, they might indicate
a short-term Byzantine control that was waning and waxing throughout
the seventh and eighth centuries. The restoration of the diocese of Stobi
that took place in the late seventh century was associated with the year 679
and the migration of Sermisianoi under Kuver from Pannonia to the
Keramisian plain (today Prilepsko polje), although Stobi had been
destroyed and left depopulated nearly a century before.529
The absence of the seventh- and eighth-century findings indicates
that the fortifications were not used in this period, as was the case with
Serbia. The high altitudes did not appeal to the Slavic tribes, which is why
the traces of their presence are to be looked for in the valleys and basins,
until the fortifications were once again re-used in the ninth and tenth
centuries, because of the war.
On the other hand, a more thorough reconnaissance of the
lowland positions was never undertaken, not in a way that would yield
adequate results. The smallest of reparations on the upper parts of the
walls are not visible today, since the relevant segments of the ramparts
have been ruined. Wooden annexes, wallings, and dwellings made of light
527
528
529
. , , 6, 1990, 45-49.
, , 32.
V. Popovi, Aux origines de la slavisation des Balkans. La constitution des
premirs Sklavinies macdoniennes vers la fin du VIme sicle, Comptes
rendus de lAcadmie des inscriptions et belles-lettres I, Paris 1980, 249-252.
186
Dejan Buli
materials have not been preserved or, in rare cases, only in fragments. It is
very common that the only indication these fortifications were occupied
are rare movable archaeological findings. Except for the jewelry and some
highly specific objects, a significant part of these findings can not be
subjected to a precise chronological determination. This applies to tools in
particular. Until recently, not enough attention was paid to the pottery, or
it was not even possible to reliably set it apart according to the epochs. This
calls for a revision and re-dating of some pottery fragments on some sites.
We will cover the ninth-century Slavic migration wave further on
in the text, when touching upon the topic in the frame of Serbia. Mikuli
holds that new brotherhoods moved from the valley of the Danube after
the collapse of the Chaganate in Pannonia (late 8th early 9th century),
when the Avars were shattered and the relations with the Slavic North
could be re-established. The contacts with the Slavic tribes beyond the
Danube have been archaeologically confirmed by numerous specific
objects, discovered at fortifications of ree (79) and Davina (65).530
During the reign of Simeon (893-927), the Bulgarian rule reached
the Drina and the Adriatic, including the entire Republic of Macedonia,
nearly touching Thessalonica. After his death, Bulgaria weakened, the
Russian prince Svyatoslav conquered it, and in 971 John Tzimiskes entered
Preslav and annexed the Bulgarian to the Byzantine Empire.
Among the standard forms of metallic findings used by the Slavic
population, occasional findings were discovered of specific objects
attributed to the Bulgarian boyars, the officers of the new administration.531
The ninth-century town of Kuprite (78) that sometimes served as a
military camp was a Proto-Bulgarian town.532 Bronze amulets representing
a horse-riding mythical hero (or a shaman) were discovered at the
fortresses of Prilep (47) and emren (5). It was assumed that these arrived
to Macedonia from the lower Danube valley, along with the expansion of
Boris and Simeons state, at the end of the ninth or in the tenth century.533
The well-known Bulgarian double-sided amulets (seals) were found at
Jegunovci near Tetovo (71); another amulet was discovered at Devol-grad
near Drenovo (29), as well as a small bronze plate with a tamgha, also dating
530
531
532
533
, , 36. Numerous toponyms such as

Morava, Boemija, Boemica, etc., corroborate this theory.
, , 35.
, , 348.
, , 83-84.
187
from the late ninth or the tenth century, from a road watch Arangel near
Kievo (31).534 The belt ornament discovered at reka (79) typologically
corresponds to the period of the collapse of the Avar khaganate. The use
of these ornaments spread to the neighbouring Slavic boyars in the early
ninth century.535
In 976, a rebellion broke out in Macedonia, led by the four
Cometopuli the sons of the comes (knez) Nikola. Samuil was the only
one to survive the uprising of 978 and he managed to place under his
authority entire Macedonia, except for Thessalonica; then he expanded his
rule to Thessaly, Western Bulgaria, Epirus, areas of Albania including
Dyrrachium and the Serbian lands, and he re-established the Patriarchate.
After the Byzantines displayed their military and technical superiority
during the campaign of 1001-1004, when they conquered Skopje and
Voden, the Empire began to crumble, persisting until the death of Ivan
Vladislav in 1018.536 A belt buckle with a representation of a griffon, a
product of Byzantine craftsmanshift, was discovered at Devol (29) and it
could be dated to this period.
After the collapse of Samuils state, Basil II was determined to
destroy the fortresses in the area that might have been used as new army
strongholds. He spared only several key castra where Byzantine military
crews were deployed. Thus were demolished Ohrid, tip and Prilep, with
only the Archdiocese of Ohrid left standing. Prosek was restored in the
late 1100s and expanded in the early thirteenth century, since it became
the centre of a new regional state.537
Of the conquered territory of Samuils state, the new theme of
Bulgaria was formed centred in Skopje, while the Archdiocese of Ohrid
was re-organized. We learn of the established ecclesiastical organization
from the Golden Bull of 1272, issued by the Byzantine emperor Michael VIII
to the Archdiocese of Ohrid. This Bull contained copies of the three Bulls
issued by Basil II to the same church in 1019, in May 1020 and between 1020
and 1025. In the Bull of 1019, 17 dioceses were listed. With each episcopal
see, towns under its jurisdiction were listed and the number of clerics and
parishioners written down. In the second Bull issued to the Archdiocese
of Ohrid, another 14 dioceses were added to the list now totalling 31
534
535
536
537
, , 84-85.
, , 85.
. , , 1959, 294-295.
, , 47.
188
Dejan Buli
dioceses.538 In this work we convey from the list of dioceses and towns only
those lying in the territory of the present-day Macedonia. The list shows
the extent of the restoration process undertaken:
1. Skopje the episcopal see (64), and the towns Bine (Serbia),
Lukovo (36), Preamor and Princip (not located) placed under its jurisdiction.
2. rovid (34), with the parishes of Kozjak (37), Slavite (35),
Zletovo (54), Pijanec and Maleevo (not located).
3. Diocese of Strumica (69) with its see most likely at the
monastery of Veljusa. Towns lying in the jurisdiction of Strumica were
Radovite (55) and Kone (68).
4. Butela Bitolj with the following towns: Prilep (47), Debret
(Deuretis) (48), Veles (77) and Pelagonija (probably Bitolj).
5. Ohrid (45) with the following towns: Kiava (32), Prespa
(Greece) and kra (lbania).
6. The area of Polog and the town of Leskovec (Leak 72) were
placed under the jurisdiction of the bishop of Prisdiana (Prizren, Serbia).
7. Prosek (40-43) and Morihovo (51), in the present-day Macedonia,
were placed under the jurisdiction of the diocese of Moglen in Greece.
Another two fortifications in the area of Ohrid Prespa were
mentioned in the treatise of John Skylitzes: one on the Prespa lake, and the
other, Vasilid, situated on a mountain top lying between the lakes of Ohrid
and Prespa.539 According to Mikuli, Vasilid was most likely one of the
two fortifications erected between the villages of Evlo and Petrino. 540 The
fortress of Termica was in the area around Strumic, and it was also
mentioned in the 1016 campaign of David Arianites.541
Archaeological findings of reliquary crosses discovered in the
vicinity of the ecclesiastical centres could easily be associated with the
establishment of the ecclesiastical organization in these areas. These sites
include Skopje, Bitolj, Ohrid, Strumica, Prilep, Prosek, Lukovica, Kozjak
and Leak. The afore-mentioned crosses, made in Byzantine workshops,
were generally attributed to the higher ranks of the tenth and eleventh
century clergy.542 An amulet with a representation of a Slavic deity, discov538
539
540
541
542
For more detailed information on this issue, including the map of dioceses
and settlements, see: . ,
, 2004, 172-177.
III, 130.
, , 275-276.
III, 119.
, , 83.
189
ered at the fortification of Davine near uer, should be dated to the times
before the Slavs in Macedonia were Christianized, rather than to the tenth
century, as the analogous findings might suggest.543
In 1040, the anti-Byzantine movement led by Petar Delyan broke
out in Belgrade and in the Morava area, because of the new taxes levied in
money and the abuse of the officials. It was not long before the revolt
spread to Ni, Skopje and Macedonia. Delyan took Dyrrachium, attacked
Thessalonica and advanced into Greece. Byzantium crushed the uprising
the same year, with the help of Alusian, son of Ivan Vladislav.544
After the crushing defeat that Byzantium suffered at Manzikert
(1071), the malcontents from Macedonia organized an uprising which
Constantine Bodin joined, after having been proclaimed emperor in
Prizren, in 1072. The rebels managed to take Skopje after defeating the
strategos of the theme of Bulgaria.545 Bodin split his armies in two groups
and headed towards Ni, while Petrilo, general of Michael VII Doukas,
took Ohrid and Devol but suffered defeat at Kastoria. Soon after, Bodin
himself was defeated in Kosovo and taken captive.546
The restoration of the Byzantine rule, along with the development
of mining in the eleventh century, had a beneficial effect on the town
growth, which culminated in the fourteenth century under the Serbian
rule. Archaeological findings discovered at fortified sites close to the mining areas indicate the renewal of the mining industry. The findings include
coins and many objects of cast iron, such as weapons and tools.547
Serbia
The social crisis that struck the Roman Empire caused striking
pauperization of the population, while the continuous flood of settlers,
various peoples and looters made the difficult situation even worse. These
groups benefited from the proximity of the frontier and the well-branched
road network to reach their loot in the flatland settlements and towns. The
Hunnic wrath caused destrucion of some important towns, such as Singidunum,
543
544
545
546
547
, , 84-85.
I, 1981, 183 (. ).
For further information regarding these events, see: . ,
- 1072 1075, 47 (2000) 35-57.
I, 1981, 190-191 (. ).
, , 50.
190
Dejan Buli
Viminatium, Margum and Naissus. It took plenty of time for these towns
to recover. The horrible times were exacerbated by the natural disasters
that befell certain parts of the Empire. The catastrophic earthquake struck
Dardania in 518,548 followed by a plague epidemic that decimated the
population and weakened the defences of the Empire.549
Insecure times called for construction of fortifications. Some of
these fortified sites were regional centres with military crews and a still
functioning ecclesiastical organization. Besides these, the imperial
authorities strived to build smaller fortifications on important strategic
points along the roads, so as to defend and oversee the communication and
supply systems. These fortified sites also served as refuge centres that
provided safe haven to the populations fleeing the endangered lowland
settlements. Parallel to the construction of these fortifications, smaller
ones were built by rural communities, to provide them with safer
positions. Although their positions changed by moving into locations on
higher altitudes, they carried on with their economic activities on earlier
agricultural fields with a shift towards pastoralism.
These measures created a new defensive system, born out of
necessity and reflecting how weak the Empire had become. The aim was
to reduce the influx of refugees that sought shelters in the south, since the
refuges were built in every part of the Empire; but also put to a heavy test
the barbarians ability to lay siege and to maintain their supply chain; in
addition, the barbarians were rather unaccomplished besiegers of
fortifications, which by then had no riches left to loot. In any event, the
smaller hordes roaming the roads of the Empire did not even pose a threat
to the villagers any longer, unless they carried out sudden attacks. But the
remains of fire on some fortifications, together with numismatic material
and relevant archaeological horizons of hoards confirm that settlements
were played havoc with, and speak of volatile times.550 This concept,
adapted for the precarious sixth century, reached its culmination during
the reign of Justinian, as was corroborated by the writings of Procopius,
but also by the plentiful material finds from throughout the Empire.
548
549
550
I, 57.
B. Geyer, Physical Factors in the Evolution of the Landscape and Land Use,
The Economic History of Byzantium. From the Seventh Through the
Fifteenth Century I (ed. A. E. Laiou), Washington 2002, 31-45.
For further information regarding the horizon of hoards from the Early
Byzantine period on the territory of Serbia, see: . - . ,
, 2006, 24-27.
The list of Early Byzantine sites on the territory of Serbia:

1. Zbradila Korbovo551
2. Rgotski kamen552
3. Gradite ukojevac, Kraljevo (9th - 11th centuries)553
4. Velika Gradina Zamanja (9th -10th centuries)554
5. Velika Gradina Miloaj, municipality of Kni (7th century)555
6. Bogut grad Bogutovac556
7. Branievo Svetinja (12th-13th centuries)557
8. Branievo Mali i Veliki grad (10th-11th,12-13th centuries)558
9. Vranjska banja Crkvite559
10. Vranjska banja Kale560
11. Gradite Korbevac561
551
552
553
554
555
556
557
558
559
560
561
. , ,
6 (1990) 115-118.
. , ,
6 (1990) 191-196.
. ,
2002-2003., 1 (2007) 39-42; . ,

, 8, 2007, 9-16. The medieval phase has
been identified by the personal insight into the material.
. , . , . . ,
,
1 (2007), 47-49.
Based on the personal insight into the unpublished material.
. , , 2/3 ..
(2004/5) 2008, 72-3.
. , . ,
38 (1988), 1-37.
. , . , ,
39 (1988), 125-176.
. , . , . . ,
,
20 (2004), 145-169 (= ,
).
. , .
1964. . , 1 (1965)
226; . , , 145-169.
. - . , ,
1951, 136; . , .
1964. . , 1
(1965) 225-226; . , , 145-169.
191
192
Dejan Buli
12. Markovo kale near Preevo562

13. Kale Klinovac563
14. Gradite Veliki Trnovac564
15. Markovo Kale Vranje (Late Middle Ages and the Ottoman period)565
16. Kacapun566
17. Stajkovac567
18. Jovac site Gradite568
19. Tesovite Otri ukar569
20. Dubnica Kitka hill (Kale)570
21. Fortifaction Sv. Ilija (15th century)571
22. Donji Romanovac Gradite572
23. Kijevac Gradite573
24. urkovica Kulite574
25. Garinje Mali Gradac575
26. Donje Balinovce Gredak (Stoliica)576
27. Mrtvica577
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
. , 1968. , 4 (1968)
502; . , , 145-169.
. -, VI ,
7 (1973), 25-37.
. , - ,
10 (1986) 59-60; , 1965, 322.
. - . , .
, 20 (1987) 141-154.
. , .
1964. . , 1 (1965)
225 (= , ).
J, , 225.
, , 225.
. . , , ,
II, 1963, 171; , , 227.
. , 1965. , 2 (1966) 322
(= , 1965.).
. - . , .
, 1977, 147-148 (= - , ).
. , 1966. 1967. , 3
(1967) 330 (= , 1966. 1967.)
, 1966. 1967, 330.
, 1966. 1967, 330.
. , 1968. , 4 (1968)
511-512 (= , 1968.).
, 1968, 512.
, 1968, 512.
193
28. Gradite Prvonek578

29. Skobalji Grad Leskovac (11th-15th centuries)579
30. Jerinin grad Gornja Crnua (16th-17th centuries)580
31. Gradina Ilinje (11th-13th,14th-15th centuries)581
32. Gradina Kaznovie (9th-10th centuries)582
33. Gradina Konuli (9th-11th, 12th centuries, Ottoman period)583
34. Gradina Lisina584
35. Kale (Grgec) Gornje Brijanje585
36. Leskovac, Hisar site (Late Middle Ages and the Ottoman period)586
37. Selite Nakrivanj587
38. Sjarina, municipality of Medvea588
39. Gornje Gradite, municipality of Lebane589
40. Kulite or Jezero590
41. Sakicol site, municipality of Lebane (the church ?)591
42. Radinovac, Gradite site (Kaljaja)592
43. Cariin Grad Lebane (10th-11th centuries)593
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
Dj. Jankovi, The Slavs in the 6th Century North Illyricum, 20

(2004) 39, 43.
. - . ,
, 27 (1987), 61-67.
. , . , . ,
-
, 16 (1988), 81-94.
. , M .
, 53 (2006), 53 91.
. , . R ,
55 (2007) 45-62.
. , , 57 (2008) 29-58.
. ,
, 54 (2007) 43-62.
. - . ,
, 1988, 39 (= -,
, ).
-, , , 41.
-, , , 41.
-, , , 37, 42;
- , , 152-153.
- , , 152.
- , , 150.
- , , 149-150.
-, , , 41.
- , ; -, ,
, 81-82; B. Bavant - V. Ivanievi, Justinijana
Prima - Cariin Grad, Beograd 2003.
194
Dejan Buli
44. Konjua Cer594

45. Nemi Donja Bukovica595
46. Vidojevica Cer (11th-12th century)596
47. Gradina Stapari597
48. Kulina near Solotua (15th century)598
49. Gradina near Bajina Bata599
50. Bregovina (10th century)600
51. Zlata601
52. Glaince Kale near itoraa602
53. Balajnac Gradite603
54. Gornji Statovac Milanov kr604
55. Bogujevac Bandera605
594
595
596
597
598
599
600
601
602
603
604
605
. Vasiljevi, M. Popovi, Konjua na Ceru . Ranovizantijsko utvrenje, AP

16 (1974) 111-112; . , . , . T ,
I, ,
1953, 43.
. , . T ,
I, , 1953, 57; . Jankovi,
Rekognosciranje srednjovekovnih nalazita u zapadnoj Srbiji i na Peteru,
AP 20 (1978) 187 (= Jankovi, Rekognosciranje).
. , . , . T ,
I, , 1953, 4546; Jankovi, Rekognosciranje, 186.
. , , ,
, 2008; . -,
. O ,
I, , 1953, 60- 61.
. - . , . O ,
I, ,
1953, 59-60.
. . ,
- . , .. 4 (2008)
143-145.
. , ,
, , - 1999, 87-116.
- , , 153; . ,
, 12-1 (1986) 217-220.
. -, . ,
1998, 54 (= -, ).
. - . , ,
9-10, 1959, 77; M. Jeremi, Balajnac. gglomration
protobyzantine fortife (Rgion de Ni, Serbie du Sud), Antiquit tardive 3
(1995) 193-207.
-, , 54.
-, , 54.
56. Rgaje Grad (10th-12th centuries)606

57. Pesti Bukoloram607
58. Miljkovica608
59. Vidovaki Kr609
60. Buince610
61. Smrdelj611
62. Babotinac Veliko Kale (the Middle Ages)612
63. Pirot (12th-14th centuries; Ottoman period)613
64. Gradite, site Grad614
65. Baranica615
66. Gradina Venac616
67. Koelj617
68. reac618
69. Kalna619
70. Sveta Trojica near Ravna620
71. Gradina Juhor (Momilov grad)621
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
-, , 44, 55, 65.

-, , 44, 55.
-, , 43.
-, , 43.
-, , 43.
-, , 43.
. -, ,
3 (1987) 213-218.
. - . , . A 1985.
, 3 (1986) 227-232; . - . ,
. A 1986. ,
4 (1987) 149-154; . , , 1996, 10-13;
. , , 8-9 (1979)
185-201.
. - . , . , 1997, 26, 113 (= - , ).
- , , 25-26.
. - . ,
, 1,
2003; . ,
, 4-5, 2011, 36-7.
- , , 27-28.
- , , 28.
- , , 29.
- , , 29.
. , , 12-1 (1986)
199-217.
195
196
Dejan Buli
72. Petrus (12th, 14th-15th centuries)622

73. Orlovia Grad Leje623
74. Mali grad Dragoevac624
75. Jerinin grad Dragoevac625
76. Jerinin grad Beoi626
77. Gradite (Devojaka stena) Sekuri627
78. Grad Opari628
79. Jerenin grad Vojska629
80. Hanite Grad, Dramirovac630
81. Brdo Kruar631
82. Bukovaka esma632
83. Govedarnik Grad, Glavinci633
84. Jerinin grad Mievi (up to the 12th century)634
85. Jerinin Grad Prevet635
86. Gradac, Banja Koviljaa636
87. Kosanin grad, Cer637
88. Trojanov grad, Cer638
89. Jerinin grad, Brangovi (9th-10th centuries)639
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
. , ,
, 2003, 281-291 (= ,
); . ,
, B 2011, 8-11; 57-66.
, , 281-291.
, , 281-291.
, , 281-291.
, , 281-291.
, , 281-291.
, , 281-291.
, , 281-291; Zaviajni muzej Jagodina.
Stalna izlobena postavka, Katalog, Jagodina 2001, 36.
, , 281-291.
, , 281-291.
, , 281-291.
, , 281-291.
, , 281-291.
, , 281-291.
. , , 2 (1985) 131
(= , ).
, , 131.
, , 131.
. - . , .
(VI-X ),
37 (2004) 79-101.
90. Gradutina, Beljin (10th-12th centuries)640

91. Grad, Lis641
92. Gradina, Vukovica642
93. Stojkovia gradina, Via643
94. Gradina, Jelica (7th-10th centuries)644
95. Sokolica, Ostra (10th-11th centuries)645
96. Lopa, site Grad646
97. Velika Gradina, Vrsenice (9th-10th centuries)647
98. Gradina, Tuzinje648
99. Jerinin grad, Trojan649
100. urevica, erekare (9th-11th centuries)650
640
641
642
643
644
645
646
647
648
649
650
. - . , .
,
I, , 1953, 16; , , 131. Ceramic findings,
dated to the period between the tenth and twelth centuries support the
hypothesis that Gradutina was used in the Middle Ages: I. Popovi, Notes
topographiques sur la rgion limitrophe entre la Pannonie Seconde et la
Msie Premire, Roman Limes on the iddle and Lower Danube,
1996, 138, note 7.
. , ,
16 (1986) 51-66; . ,
2002. 2006. , 36 (2006) 31-48;
. , . , 4 (..) (2008) 146-150.
. , . ,
, 16 (986) 75-80.
. , . , . ,
, 16 (1986)
67-74.
. , .
, 2010. For further information on the
medieval strata, see: . ,
, 50 (2004) 153-204.
. , ,
2, 1995, 53-58; . ,
X-XI , 19
(2003) 223-247.
, 2011, 39-40.
M. Popovi - V. Biki, Vrsenice. asnoantiko i srpsko ranosrednjovekovno
utvrenje, Beograd 2009 (= Popovi - Biki, Vrsenice).
. -,
, 6 (1982) 242-243.
. , , 13 (1989) 7-15.
. , ,
7 (1983) 29-37. We have established the medieval stratum ourselves,
based on the published supplemental table with the ceramics.
197
198
Dejan Buli
101. Gradina, Hum651

102. Gradina, Ramoevo652
103. arski kr, Duga Poljana653
104. Gradovi, aronje654
105. Gradina, Radalica655
106. Kulina, Rogatac656
107. Zlostup, Ostrovica657
108. Litice, Dobrinja658
109. Juac, Sopoani659
110. Gradina, aronje660
111. Pazarite Novi Pazar (9th- 11th, 11th-13th centuries)661
112. Izbeg, Tupi kr662
113. Kula, Kaludra663
114. Gaj, Babre664
115. Grad (Gradina), Nosoljin665
116. Gradina, Postenje (9th-12th centuries)666
117. Kr, Zlatni Kamen667
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
. , ,
12 (1988) 5-11.
. , , 11 (1987) 5-11.
. , ,
7 (1983) 5-14.
. ,
, 6 (1982) 131-140.
. . , , 9 (1985) 39-46.
. , , 11 (1987) 13-20.
. ,
, 9 (1985) 47-54.
. , . , 6 (1982) 238-239.
. Popovi, Juac kod Sopoana, AP (za 1986. godinu), Ljubljana 1987, 115-117.
. . , ,
3, 1988, 51-68
. Popovi, Tvrava Ras, Beograd 1999.
. ,
, 9 (1985) 47-54.
. , , 8 (1984) 11-18.
. -, , 13
(1989) 17-27.
. , . , , 7 (1983) 21-27.
. , -.
1994. , 11 (1996) 198-207; . , . a 1995. , 12 (1997) 121-129.
. ,
, 14 (1990) 7-17.
118. Panojevii668
119. Matovii669
120. uprija, medieval Ravna (ruined in 1183)670
121. Slatina, near Brza Palanka (8th-10th centuries)671
122. Mokranje (Petres) (11th century)672
123. Mihajlovac Blato (necropolis, 10th century)673
124. Kula Mihajlovac (7th, 9th-10th centuries)674
125. Majur (Jagodina) (7th century)675
126. Kostol Trajanov most (Pontes) (10th-12th centuries)676
127. Korbovo (7th century; the entire Middle Ages)677
128. Gamzigrad (Romuliana) (11th century)678
129. Prahovo (Aquae - Akvis) (7th, 9th-11th centuries)679
668
669
670
671
672
673
674
675
676
677
678
679
Popovi - Biki, Vrsenice, 126.

Popovi - Biki, Vrsenice, 126.
. . , , 1990,
115 (= . . , ).
. , , II
(1984) 196-198; . , . , . , Lembouchure de la
rivire Slatinska reka, III (1986) 378-387; . . , ,
110.
. . , Mokranje kod Negotina. Kamenolom - vieslojni lokalitet,
AP 18 (1976) 22-24; . , .
, II (1984) 221-225; . . , , 103.
. - . , Mihajlovac. ntiko utvrenje, AP 6, 1964;
. , Mihajlovac Blato. Une forteresse de la basse antiquit,
III (1986) 404; . . , , 101.
. , Le site dhabitation mdival kula prs du village Mihajlovac,
III (1986) 443-446; . . , , 101-103.
. . , , 100.
. - . , . K Pontes, I (1980) 2324; . - . , Castrum Pontes, IV (1987) 81; .
- . -, . Castrum Pontes,
II (1984) 44-47; . -, Pontes. .
, IV (1987) 135-136; . -,

Pontes, IV (1987) 117-119.
. , VI VII
, 1981, 194 (= , ); . ,

IX-XI , 11 (1983) 101; . . , , 95.
Byzantine authorities renewed the town in the early eleventh century and
its population was resettled after the uprising of 1072: . . ,
, 87-89.
, , 43-45.
199
200
Dejan Buli
130. Saldum680
131. Bosman681
132. Kulina Medvednik682
133. Mora Vagei683
134. Borej684
135. Tekija (Transdierna?) (10th-11th centuries)685
136. Karata (Diana)686
137. Donje Butorke687
138. Glamija Rtkovo688
139. Vajuga Karaula (medieval necropolis)689
140. Milutinovac690
141. Ljubievac691
142. Radujevac Karamizar692
143. Site at the mouth of river Timok693
680
681
682
683
684
685
686
687
688
689
690
691
692
693
. , Saldum. o o
, 33-34 (1982/83) 319-331.
. , . , 33-34
(1982/83) 137-144
.
. - - . , La forteresse antique Mora
Vagei prs de Mihajlovac (Fouiles de 1981), 3 (1986) 453-466; P. pehar,
Materijalna kultura iz ranovizantijskih utvrenja u erdapu, Beograd 2010,
44-45 (= pehar, Materijalna kultura).
. -, . , . K
, 2 (1984) 217-220.
. Cermanovi-Kuzmanovi, A. Jovanovi, Tekija, Belgrade 2004; . .
, , 114.
. , .
, , ,
2006, 115-122, including the bibliography.
, , 35; pehar, Materijalna kultura, 30-31.
, , 41; . , . (
1980. ), 2 (1984) 109; pehar,
Materijalna kultura, 35.
, , 41; pehar, Materijalna kultura, 35-37 .
, , 45; . Kora, Late Roman and Early Byzantine
Fort of Ljubievac, Roman Limes on the Middle and Lower Danube (ed: P.
Petrovi) Belgrade 1996, 105-110; pehar, Materijalna kultura, 48.
, , 45; . - . ,
, 8 (1991) 144-151.
144. Vinjica (Octavum)694

145. Jerinin Grad Gradac by Batoina (medieval necropolis)695
146. Vidrovgrad Vidrovac696
147. Gradina Veljkovo697
148. Tabakovako Brdo698
149. Gradite Gradskovo699
150. Gradina Grbice (10th-11th centuries)700
151. Gola Stena tubik701
152. Brza palanka (Egeta)702
153. Sokolica703
154. uka Podrka704
155. Miro705
156. Sip706
157. Trajanov Most 2 Kostol (12th century)707
158. Sirmium (the entire Middle Ages and the Ottoman period)708
159. Basiana709
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
M. Birtaevi, Vinjica - Castrum Octavum; Beograd. Vizantijsko utvrenje

i nekropola, AP 6 (1964) 109-111.
. , (
), 13-14, 1965, 275-290 ; . ,
, 1981, 56.
, , 49.
, , 49-50.
, , 51.
, , 51.
. , ,
1981, 57-8; . , ,
, 1985, 26.
, , 52.
. , - , II (1984) 153-166; P. Petrovi,
Brza Palanka - Egeta, III (1986) 369-377.
, , 54.
, , 54.
, , 54-56.
, , 56-58; . ,
1072. , 25 (1978) 52.
. , . (
), , 2003; . .
, , 114; , , 281284 (. ).
Arheoloki leksikon (ured. D. Srejovi), Beograd 1997, 112 (A. Jovanovi) (=
Arheoloki leksikon); M. orevi, Arheoloka nalazita rimskog perioda u
Vojvodini, Beograd 2007, 45-49.
201
202
Dejan Buli
160. Remesiana710
161. Beograd (Singidunum) (9th-15th centuries; Ottoman period)711
162. Dubravica (Margum) (10th-11th centuries)712
163. Ram (Lederata) (10th-11th centuries)713
164. Veliko Gradite (Pincum)714
165. Golubac (Cuppae)715
166. Boljetin (Smorna) (9th, 12th-15th centuries)716
167. Ravna (Campsa) (9th-11th centuries; necropolis, 14th-15th centuries)717
168. Poreka reka718
169. Sapaja (12th century; Turkish and Austrian period)719
710
711
712
713
714
715
716
717
718
719
. , , 1999, 101-110.
For now, the earliest traces of material culture in Belgrade came from the
slopes along the river Sava (the Lower Town and the Western suburb), and
date back to the ninth or, possibly, tenth century: . -,
, 25 (1978) 7-16; ,
, 31-43 (. ).
. , 1989.
. , 14, 73-75; . ,
1990. ,
15, 39-40; D. Spasi-uri, Die rmische Stadt-Margum, Margum,
Poarevac 2003, 11-24; Arheoloki leksikon, 630-631 (A. Jovanovi); .
, 1072. ,
25 (1978) 41-55 .
Arheoloki leksikon, 576-577 (A. Jovanovi); A. Jovanovi, The Problem of
the Location of Lederata, Roman Limes on the Middle and Lower Danube
(ed. P. Petrovi), Belgrade 1996, 69-72; . . ,
1072. , 25 (1978), 41-55.
M. Mirkovi, Rimski gradovi na Dunavu u Gornjoj Meziji, Beograd 1968,
101-103; Arheoloki leksikon, 811 (A. Jovanovi).
. , 1072.
, 25 (1978) 43; Arheoloki leksikon, 555-556 (A. Jovanovi).
. , (Smorna). ,
33-34 (1982/83) 211-225; . -, .
, 33-34 (1982/83) 227-230.
. , (Campsa). o o
33-34 (1982/83) 233-251; . -, .
, 33-34 (1982/83) 253-257.
. , .
, 33-34 (1982/83) 285-291; pehar, Materijalna
kultura, 20-22.
. , .
, 33-34 (1982/83) 29-62; . ,
,
1995.
203
170. Veliki Gradac (Taliata) (7th, 11th-12th centuries)720

171. Hajduka Vodenica (11th-15th centuries)721
172. Malo Golubinje722
173. Veliko Golubinje723
174. ezava (Castrum Novae) (necropolis 11th-13th centuries; tomb 17th century)724
175. Ni (Naissus) (11th-15th centuries; Turkish period)725
176. Medijana (intermittent settling in the Middle Ages)726
177. Pajkovac Gradite727
178. Bedem Maskare728
179. Veliki Vetren729
180. Donji Dubi Gradite730
181. Puhovac Gradite731
182. Laisled Gradite732
183. Ukosa (Kuite) Stala (10th-11th centuries)733
720
721
722
723
724
725
726
727
728
729
730
731
732
733
. , - (Taliata).
, 33-34 (1982/83) 265-282; .
, -I ,
1981.
. , .
, 33-34 (1982/83) 319-331; . -,
. , 3334 (1982/83) 333-336.
. , . ,
33-34 (1982/83) 297-300 (= , ).
, , 297-300.
. , - Castrum Novae, 33-34 (1982/83) 319-331.
. , , 1999; . ,
, 31 (1984) 5-40.
. , , 2006.
. ,
, 13 (2008) 9-52 (= ,
).
, , 9-52.
, , 9-52.
. . , ,
18 (2002) 137-156 (= , ).
, , 150-152.
, , 153-154.
. , (), 3,
1988, 281-282; ,
2000, 18 (= ); . ,

, 3, 2005, 187-188.
204
Dejan Buli
184. Ljubinci Gradite (9th-11th centuries)734

185. Plonik Gradac735
186. Vitkovac Gradac736
187. Boljevac ukar737
188. Porodin Gradite738
189. Petina Gradac (13-14. centuries)739
190. Jablanica daje740
191. Zlatari Gradite741
192. Gradac erekari in Gornji Levii (9th-10th centuries)742
193. Dupci Gradite743
194. Brus Gobelja Gradite (9th-11th centuries)744
195. Gradite Trnavci745
196. Koznik746
734
735
736
737
738
739
740
741
742
743
744
745
746
. , . ,
9 (1993) 228-235; , 17.
, 17.
, 17; . - . ,
,
7, 2009, 188.
, 21; . ,
,
3, 2005, 185 -186.
, 21.
, 24.
, 25.
, 24.
, 28; . - . ,
,
44/1 (2007) 27 45; . ,
, 3,
2005, 189.
. - . ,
, 7,
2009, 186, 188, . 5/4 7/2; , 26.
. -, . ,
, 32 33 (2002) 99-120.
. - . ,
, 44/1 (2007) 38.
. , . , , 2, 1980, 307;
, , 218-222 (. ). We were told of the
existence of the Early Byzantine layer, from the unpublished excavations
undertaken by . Jankovi.
197. Kaljaja Grgure near Blace747

198. Fortification in Pridvorica, Kurvingrad748
199. Gradina Dedinci749
200. Duvarine Via (the town of Toplica Milan) (Middle Ages)750
201. Jelika uka, Saganjevo751
202. Gradina Donja Rudnica, near Raka752
203. Viegrad (11th-12th centuries)753
204. Gradite Gee (9th-10th centuries)754
205. Radavac in the vicinity of Pe755
206. Jerinin grad Dolac (Late Middle Ages and Turkish period)756
207. Crmljani near akovica (Middle Ages)757
208. Ere758
209. Kusare759
210. Gradite Zatri760
211. ablanica near Pe761
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
. ,
,
, - 2002, 64.
Based on personal insight.
. , ,
13 (1997) 147-158.
. , XI-XV , .
, - 1987, 371 (=
, ); . , . , 11 (1995)
208-223.
, , 371; Luan Przhita, Gezem Hoxha,
Fortifikime te sheujve IV-VI n Dardanin Perndimore, Tiran 2003, 143
(= Przhita, Hoxha, Fortifikime);
, 2002 (= ), 98.
Przhita, Hoxha, Fortifikime, 145.
, , 375; Przhita, Hoxha, Fortifikime, 145-146.
, 93; Przhita, Hoxha, Fortifikime, 146.
This site was drawn on the map, among fortifications from the period
between the fourth and sixth centuries, but was not mentioned in the text:
Przhita, Hoxha, Fortifikime, 66 .
Przhita, Hoxha, Fortifikime, 146-147.
, 373; Przhita, Hoxha, Fortifikime, 147.
205
206
Dejan Buli
212. Dobre Vode near Klina (Late Middle Ages)762

213. Kaljaja Orahovac763
214. Gradina Vrani (tombs bearing similarities to the Komani-Kruje culture)760
215. Hisar Kostrc near Suva reka (11th-15th centuries)765
216. Gradina Koria (Middle Ages)766
217. erinina kula Podgrae near Klina767
218. Gradina uti kamen (Guri i Kuq)768
219. ean (10th century; 14th century)769
220. Zvean (throughout Middle Ages)770
221. Gradina Gornji Streoci771
222. Gradina Crni vrh772
223. ilivode773
224. Ulpiana Lipljan (Iustiniana secunda)774
225. Gradina Drsnik775
762
763
764
765
766
767
768
769
770
771
772
773
774
775
, , 372.
S. Fidanovski, Kostrc. neolitsko naselje i ranovizantijsko utvrenje, AP
(1986), Ljubljana 1987, 48-49; , , 375; .
, ,
, 1998, 278
(= , ); . , ,

, 1998, 370; Przhita, Hoxha, Fortifikime, 149-150.
, , 372; Przhita, Hoxha, Fortifikime, 147-148.
Przhita, Hoxha, Fortifikime, 152.
V. Ivanievi, P. pehar, Early Byzantine Finds from ean and Gornji
Streoc, 55 (2005) 133-159 (= Ivanievi, pehar, Early Byzantine).
. , ,
, 1950, 169-170; , , 373375; . ,
, , 1988, 35-36 (= , );
, , 112-115 (. ).
, , 371; V. Ivanievi, P. pehar, Early
Byzantine Finds from ean and Gornji Streoc, 55 (2005) 133159 (= Ivanievi, pehar, Early Byzantine).
, , 371; Ivanievi, pehar, Early Byzantine, 159.
Ivanievi, pehar, Early Byzantine, 159.
. -,
, 32 (1981) 57-75; ,
, 376; , , 342-344.
, 64.
226. Veletin (10th-12th centuries; 14th-15th centuries)776

227. Zidanac near Gotovua777
228. Rimsko gradite-Brezovica778
229. Mali Petri (14th century)779
230. Veliki Petri (14th century)780
231. Kulina Tenedol (Late Middle Ages)781
232. Stanior Prekopite782
233. Gradina uaica in Guvnite783
234. Gradite Trpeza (Late Middle Ages)784
235. Gradite Grnar785
236. Kaljaja (Gradite) Bina (Late Middle Ages)786
237. Markov kamen Topilo (Middle Ages)787
238. Gradina Ariljaa (Middle Ages)788
239. Banjica Vuak (Middle Ages)789
240. Kaljaja Vrbovac (Middle Ages)790
241. Koretite Granarica791
242. Jerinin Grad Tolisavac792
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
E. Shukriu, Veletin, Multistrata Settlement, AP 29 (1988) 1990, 104-106; .

, , 53-54 (20003/2004) 139-161.
. , .
, 10 (1997)
31-35 (= , ).
, , 31-35.
. , - ( ), 1 (1950)
219-222; , , 368-369; ,
, 218-222 (. ). The Early Byzantine layer was established
to exist according to the information provided by . Jankovi.
According to . Jankovi, the existence of the Early Byzantine layer was confirmed.
H. Mehnetaj, Kulina a Vogl (Kulina Tenedol), Vendobanim
shumshtresor (Multistrata Settlement), AP 29 (1988) 1990, 96-99.
, , 278.
. - . ,
, 44/1 (2007) 39, 43.
, 384; , 142.
, 371; , 142.
, 371; , 142.
, 383; , 141.
, 367; , 141.
, 140.
, 140.
, 90.
. , -, 1998, 17 (= , ); .
, . , ,
I, , 1953, 48.
207
208
Dejan Buli
243. Kostajnik793
244. Gradac Dvorska794
245. Gradac Vrhpolje795
246. Gradina on the Orovika mountain796
247. Gradina Mikuljak797
248. Gradina Pridvorica798
249. Zasad Petrova799
250. Gradite Osladi800
251. Gradina engolj801
252. Gradina Drenik802
253. Gradina Ravni803
254. Gradina Mokra Gora804
255. Gradina Svrakovo805
256. Gradina Radobua806
257. Gradina Visoka (Golubinjak)807
258. Gradina Kruica808
259. Gradina Krstac809
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
, , 17; . , . , .
, I, ,
1953, 49.
, , 17.
, , 17; M. Vasiljevi, Arheoloko rekognosciranje Podrinja,
AP 18, 171.
, , 17; . , . ,
I, , 1953, 57-8.
, , 17; . , . ,
I, ,
1953, 58.
. ,
, 6 (1990) 208-9 (= ,
).
, , 211.
. , (III-IV ), (
1918) I, 1989, 126-127.
Based on personally carried out reconnaissance.
. , ,
1, 2001, 35.
209
Most of the afore-listed sites were registered by reconnaissance or

sondages and on fewest fortifications were conducted systematic
excavations. Some random information was obtained by the study of
material the material unearthed in the illegal excavations by looters
euphemistically called "amateur archaeologists".
According to the compiled list, 259 fortifications used in the
period ranging from the fifth to the early seventh century have been
registered in Serbia up to the present. But this number is only temporarily
correct, as it is constantly getting higher because of the fieldwork
dynamics in some parts of Serbia.
The majority of these fortifications is located on high altitudes,
often on locations difficult to reach. The wide area of the present-day
Mava was left almost entirely depopulated and without forts; first
fortified sites were erected on the mountain slopes of Rudnik, Cer and
Majevica. Lowland fortifications were primarily situated along the limes and
built mainly for military purposes; but they were now inhabited by both
soldiers and civilians. If this is the case, it is an example of either continuous
use of the fortifications or of their restoration. Only the crews in fortresses
and towns were made up of actual army, whereas fortified settlements were
defended by their own denizens. The supply was carried out with ships and
the presence of amphorae is a sure sign of military presence.
A large number of the examined Early Byzantine fortifications was
single-layered, that is to say that in most of the cases, they were re-built
during Justinians Restoration. In some areas, the restoration process was
predominant, as shown on the example of Dardania, which does not
necessarily mean that the rebuilt fortifications were built considerably
before the early sixth century. Several fortifications have never been
restored, after having suffered destruction in the first half of the sixth
century such as Kale in the village of Klinovce (13), destroyed during the
Kutrigur incursion.810
In Serbia, the lowland towns situated along the Danubian limes
and with urban tradition, were pillars of defence. On the other hand, the
newly-founded regional centres - of which the most researched are
Cariin grad (45) and Gradina on Jelica (94) - display the utmost potential
810
Money deposits could be related to the Kutrigurian incursion of 544: . , VI

, VII (1973)
25-37.
210
Dejan Buli
of the Early Byzantine construction and the urbanistic notions of the age,
in the Balkans. Some of these fortifications were built on lower, more
accessible grounds, on strategically important points that secured traffic
ways or supply routes or protecting ports as was the case with BedemMaskare (178). Among these are the fortifications, such as Gamzigrad
(128) and Mediana (176), of specific purpose or erected on the foundations
of ancient Roman palaces.
Churches existed at a large number of sites and other buildings, as
were multiple layers containing various movable findings. In most of the
high fortifications, the assortment of these findings indicates the presence
of a civilian population, refuting the hypothesis that these were refuges,
and indicating that these were more likely fortified villages involved in
mining and the communication-system control. Certainly, a smaller
number could have been refuges. These fortifications represented the basic
settlement-unit of the Illyrian provinces and could have been nothing
more than rural settlements, i.e. villages,811 until the circulation of money
finally ceased in 615, and with it monetary trade and presence of the
state.812 A long gap followed before these fortified sites would be used
again, apart from some rare exceptions.
Beside the Romaion population and other subjects of the Empire,
there were other ethnical groups living in the fortifications: Germanic
peoples primarily, but also individuals of nomadic and Slavic origin. In all
likelihood, they came there after the confrontation of the Avars and the
Langobards with the Gepids, in 567. The Empire was trying to solve the
chronic lack of manpower, caused by the Hunnic scourge in the fifth
century and the recurring Avaro-Slavic incursions of the sixth century. The
depopulation was exacerbated by a great plague epidemic and an earthquake.
With all the devastation and havoc caused by the permanent raids, the
ever-present danger and insecurity, the population fled their homes and
retreated towards the coastal towns and the safer provinces of the Empire.
The Slavs joined the ranks of the Byzantine army as individuals
and fought in wars in Italy and Asia Minor, where some of them were
promoted to officers (Hilwud). After the Avar conquests in 584 586,
some of the fortresses remained derelict and Byzantium left the defence of
811
812
. , .
, 2010, 228.
V. Popovi, Les temoins archologiques des invasions avaro-slaves dans
lIllyricum byzantin, MEFRA 87, Rome 1975, 494-496, 502-504.
211
some fortresses to the Slavs, as was the case with Alicaniburgo. In spite of
its abandoned ramparts, a Slavic settlement Dunav (Slatina) (121) sprang
up just before the end of the sixth century.813 Archaeological findings from
Gamzigrad (128)814 and the rural settlement Reka Vina, near
Belgrade815 confirm a mutually consensual Slavic colonization. The
fortification situated on the hill Govedarnik above Majur, near Jagodina
(125) should be included among the Early Byzantine fortifications
containing Slavic findings. Accidental findings include a bronze fibula
dated to the early seventh century, a ring made of lead, and a small cross,
most probably from the seventh century.816
Findings of medieval pottery were rare in Early Byzantine
fortifications, but not as rare as was thought at first. But now, this view is
changing. Experience gained over time allowed for an easier distinction to
be made between potteries from the two epochs, which was not possible
initially. Because of this distinction, today we can, in some cases, speak of
a medieval presence and that the percentage of such sites keeps rising.
Rare seventh-century pottery fragments were discovered in a series of
fortifications: Veliki Gradac (170); ekija (135), Gradina-elica (94), Slatina
(121), Kula-Mihajlovac (124) and Velika Gradina at Miloaji (5). The tomb
discovered at Kamenovo near Petrovac on Mlava was also dated to the
early seventh century, while sporadic findings of fibulae unearthed in
Prahovo (129) and Korbovo (127) point toward the existence of tombs
containing female skeletons.817
It is assumed that with the fall of the limes, Byzantine hold did not
fully disappear, because some accidental findings indicate Byzantine
813
814
815
816
817
. . , , 18.
Within the Early Byzantine layer at Gamzigrad a house was discovered that
contained Slavic objects from the period ending with 584/6: . .
, , 87.
. . , , 82-84; . , , .
, 25 (1986), 61-63; . ,
( ) 27 (1985) 120; . ,
. VI , 36 (1990)
5-16. Four sunken huts were discovered, together with the artisnanal objects
and a grave, dated between the sixth and the early eleventh centuries. Dating
the settlement to the sixth century was carried out with the Byzantine
products, primarily pottery. The only grave, of a female person in a fetal
position, was also dated to the sixth century. The sunken hut with a stone
oven is from the early seventh century, i.e. from before the reign of Heraclius.
. . , , 100.
. . , , 25.
212
Dejan Buli
presence in the Danubian basin, which would match the assumed role of
the Serbs and the Croats as foederati. Buckles from the seventh century,
discovered in Prahovo (129), Kostol (126) and uprija (120), coins of
Constantine IV, unearthed near Jagodina and dated to 634/4, and the
Byzantine wheel-made pottery discovered in Dunav (121) and Kula (124)
- confirm that hypothesis.818 It would appear that in the seventh century
Byzantium still held strongholds along the Danube and along the road
Mitrovica-Belgrade-Ni-Sofia-Constantinople. This situation changed
only with the Bulgarian incursion in 680.819 The often disputed remark of
Constantine Porphyrogenitos that a Byzantine strategos was present in
Belgrade at the time of the Serb arrival, implies that Byzantium did
manage to preserve some form of authority over the northern Illyricum,
even after 614/5.820
After their arrival, the Slavs encountered two types of settlements.
For one, towns from the Antiquity underwent significant reconstructions
in the sixth century, accordant with the Byzantine construction policy.
The answer to the question whether the Slavs immediately occupied the
fortifications, is to be found in the ethnic attribution of fragments of handmade pottery discovered on the sites. The dilemma has not yet been solved
if the ceramics are Slavic, in that case present at the beginning of the
seventh century, or if it was made by the autochthonous population, who
had to rely on the local production of ware once the trade stopped.
Interpreting several forms that seem to replicate Early Byzantine pottery
forms, purports the latter hypothesis, especially since no recognizable
Slavic pottery of a later date has been found in the areas of the sites where
the above-mentioned pottery of Byzantine form was discovered. But
before any ethnic attribution is made, it should be well considered if these
vessels may have had a special function, such as metal-casting, which can
be confirmed by the analysis of the interior. Hand-made vessels made for
this function were discovered in Duklja (Doclea).821
The absence of storage ceramics and luxury objects corresponds
with commerce, craftsmanship and money circulation becoming defunct.
But pottery production, especially of cookware, is a local activity and a
818
819
820
821
. . , , 19.
. . , , 20.
- , , 180.
D. Drakovi - M. ivanovi, Keramika prostorije 3/IX. Prilog poznavanju
svakodnevnog ivota antike Duklje, Nova antika Duklja II, Podgorica
2011, 76-77.
213
complete halt of its production would be unlikely. Remains of fire places

above the layer of destruction at the basilica D offer some insight. They
suggest the use of the church for habitation even after its demise in a fire
evident from many traces of soot. Many fragments of characteristic Early
Byzantine ceramics were discovered in the context of the fire places, giving
basis to the assumption that there was no significant time gap between the
creation of the fire places and the ruin of the church.822 This fact sets a good
example of how a local population, even after catastrophes, can continue to
use locally produced, but distinctive wheel-made ceramics, while the use of
hand-made ceramics can be attributed to another ethnos, and not explained
away with a simple early seventh-century ruralization. Also, considering
the disregard for the sacral place, we can assume that another ethnos used
the existing material culture, which can give us an answer of why there are
so few traces of Slavic material culture and this would not be an isolated
case, since findings of Romaion ceramics appear in the Slavic settlement of
Kula-Mihajlovac as well (124).
Setting this issue aside, no matter if the ceramics is Slavic or made
by the autochthonous populations, it is from the seventh century and after
it comes a hiatus until the ninth century on thee fortifications. The
exceptions are very rare and still unexplained. To be more precise, for now
we know of only one such high site in Serbia Gradina on Jelica (94).823
The shortcoming of researches conducted until now in Serbia is the lack of
reliable methods that would allow for precise chronological datings of the
discovered findings above all, the C14 method. A lay person might
conclude that maybe among the unearthed findings of pottery there are
some that could be dated further back into the past. Argumentation in
favour of this deliberation is reduced to pottery, since other types of
findings are quite rare. On the one hand, the characteristics of the
typology and style of the findings (tools, weapons) do not allow for precise
dating. Moreover, such findings are rarer than findings of pottery and
independent discoveries of such findings do not allow placing them into
specific epochs.824
822
823
824
, , 180.
. ,
, 50 (2004) 153-204. Slatina (121) is a lowland site,
which lasted in a continuum until the ninth century.
So an apsurd situation happens that from a great site that has been
systematically excavated for years, we have almost no object, weapon or
tool that we could unequivocally declare medieval except for the many
findings of pottery, and some buildings: , .
214
Dejan Buli
Traces of fire indicate that a certain number of these fortifications

perished in fires and show that life ended in a violent manner. This
destruction was caused by the Avaro-Slavic incursions after which began
the Slavic colonization of the areas south of the Sava and the Danube,
when civilized life was discontinued. The Serbs and the Croats, along with
other Slavic groups, would become in the centuries to follow the regions
prevailing population.
Life in Singidunum was abruptly brought to an end, or the town
had already been destroyed and ravaged. The new name of Belgrade speaks
clearly of the discontinuity. Belgrade was mentioned as a diocese in 878,
meaning it was an important centre, possibly since Krum brought the
middle-Danube lands under his rule.825
Is this case paradigmatic? Did towns such as Belgrade, vanguards
exposed to assaults on the Empires frontiers, face discontinuity because of
their disadvantageous geographic position? Arguments support this
hypothesis. The situation was similar in other Early Byzantine
fortifications on the Danube. The earliest agrarian settlements develop in
the ninth and tenth centuries, except on the sites of Slatina (121) and
Aquae (129).
In the territory of Serbia, the most northern fortifications such
as Brangovii (89), point to the ninth century as the earliest medieval
phase, thus reinforcing the hypothesis that other fortifications were not
occupied in the period between the arrival of the Slavs and the ninth
century. Current mental image has been based on the current level of
research, which is not quite exemplary, but which shows that the Slavs
most often settled by a river on flat or slightly hilly landscapes. The
question is posed where are the settlements and necropoles that should
exist? South of the Sava and the Danube, there were just a few of such
locations. The reasons are rather banal. Focussed research in this
direction and on these localities simply never took place. It has been
simpler to register Slavic presence by researching already existing
Byzantine fortifications, rather than to obtain these results by planned
research.826
825
826
. , , .
, 2006, 27-29.
Such an endeavour would demand systematic and organized reconnaissance
of the locations in the flatlands and river valleys, followed by systematic and
expensive excavations of these sites; and for something of that scale there
was never any money, nor political interest.
215
The score of registered fortifications is high, and it is increasing

every day, because there are still areas in the present-day Serbia where
detailed reconnaissance was not conducted. Expressions for remains of
fortifications appear as toponyms and point to these fortifications, especially
if they are of a more recent date: Gradina, Gradite, Gradac, Gradaac,
Graica, Kula, Jerinin Grad. These are mainly pre-historic, Roman and Early
Byzantine fortifications that had been there, but also the ones that through
time lost their names, including the ones with a strong medieval phase.827
Slavic names gradina, gradite, etc. do not reveal a Slavic settlement, but
indicate a Slavic settlement nearby, a Slavic environment, so to speak. The
term gradina also designated a medieval town that lost its name in time for
whatever reason, even at a much later date.828
This loss of the name also shows that of the gradinas occupied in
the Middle Ages (which was not a small number), only those that were
occupied in the Late Middle Ages as well, preserved their name. Those
that were occupied until about the twelfth century, usually lost their
names. The discovered material points to temporary residences, without
major economic activity or some important functions, even for the local
people except in the need for haven. As places irrelevant in economic
sense, in times of peace they were quickly forgotten unless they gained a
more important role in the Late Middle Ages, when their names were
preserved. Frequent movements of population reduced the appellation to
a general term for such kind of structures, visited only by pastoral
populations. Also, many Late Medieval fortifications did not retain their
names, probably because they were used for a short time.
Considering the inter-relation of the terms Grad-Gradina-GraditeGradac, etc. (living settlement, abandoned fortification, large/small fortification, destroyed fortification), one should keep in mind the lack of a clear
distinction between the terms and that they are synonymous with an
occasional particular meaning in local dialects.
The nature of the contact with the autochthonous population can
only be speculated on; but the adoptions of the toponyms, of the local
sacral places and even of the objects from the material culture, confirm
that this contact took place. At Gradina on Jelica, the Slavic population
827
828
For the meaning of the mentioned terms, see: . ,

, . , 2006, 31.
. , , .
, 2006, 31-32.
216
Dejan Buli
adopted from the autochthonous population the way of food-preparation

on a hearth by using a shallow-bell lid (sa, vrnik), which mimicked the
Early Byzantine type of the cookware, but with a different decoration.829
Other linguistic events from the history of the Serbian language also
witness of the meeting between two different linguistic and cultural
environments.830 The adoption of some of the river names also speaks that
there were mutual contacts and a certain continuity. Indirectly, it also
points towards the spaces the Slavs settled at first, fertile river valleys. One
of the examples is the name of Ras. When Procopius of Caesarea portrayed
the construction activity of Emperor Justinian, he marked Arsa among the
fortifications in Dardania.831 Constantine Porphyrogenitos was the first to
mention the medieval Slavic form of this name: the form Rasa derived
from the pre-Slavic name Arsa, a change that could only have happened
as a product of metathesis of liquid consonants, a well-known and familiar
linguistic phenomenon in Europe. This phenomenon occurred during the
first years of life of the Slavic settlers in the new environment. In the
Balkans, this process concluded in the ninth century.832 Among known
analogies from the wider area of Slavic settlement are: Arsia-Raa, a river
in Istria, Arba-Rab, an island in Croatia, and Albona-Labin.833
* * *
We will attempt to shed light on the medieval events in certain
fortifications by looking at them through the historical context. If
Constantine Porphyrogenitos is to be believed, the Serbs began to settle
the Balkan Peninsula during the reign of emperor Heraclius.834 The Serbs
took the most of Dalmatia, i.e. the territories of the present-day Serbia, of
829
830
831
832
833
834
. ,
, 50 (2004) 153-204.
. , , 1971, 23-24;
I, 1981, 128-129, 131 (. ).
I, 61.
. , , . , 2006,
16-17.
P. Skok, Slavenstvo i romanstvo na Jadranskim otocima I, Zagreb 1950, 57.
Constantine Porphyrogenitus, De administrando imperio I (ed. Gy.
Moravcsik R. J. H. Jenkins), Washington 1967, 32.7-12 (= DAI);
II (. . ),
1959, 47 (= II).
217
Pagania, Zachlumia, Travunia and Konavle,835 as Einhard mentioned in the

episode of Liudewitus escape.836 After the Croats and the Serbs had settled,
historical accounts made no further mention of Avar incursions.837
Conversion of the Serbs into Christianity took place during the reign of
Heraclian dynasty, but the decisive point happened in the mid-ninth
century, when Christian names were, for the first time, given to the
children of Serbian rulers.838 There are assumptions that earlier, individual
conversions to Christianity happened during the sixth century.839
Events in Serbia were closely intertwined with relations between
Byzantium and Bulgaria. The first Serbo-Bulgarian conflict happened in
848, during the reign of the Bulgarian khan Presiam and Serbian archont
Vlastimir, and it lasted for three years, until 851; soon after, in 853/4, the
second war broke out.840 In the mid-ninth century, the border area was
around Ras, be it a town or an area, which is why an entire string of
fortifications has medieval layers from this period. The prevailing opinion
holds that the Slavs, as an agrarian population, settled along river valleys,
in fields, beyond urban units. It was only during the first Serbo-Bulgarian
war that the fortified sites were used again, serving as refuges and
important military strongholds roles they would play in later conflicts
with the Bulgarians.841 Several fortifications might have been used in the
conflicts between members of the ruling family, as well.
835
836
837
838
839
840
841
DAI I, 32.21-25; II, 49.

Einhardi Annales (ed. G. Pertz), MGH SS I, Hannoverae 1826, 209.13-17.
According to ivkovi, the Serbs arrived as foederati between 630 and 634,
to prevent Avar incursions. For more extensive information on this issue,
including the relevant bibliography, see: , , 271291. However, the idea that, before the Serbs arrived to the Balkans with
the imperial concession and with the benefits coming from their status as
foederati, such a status (of foederati) had been bestowed upon certain
groups of the Slavs and the Antes since the reign of Justinian, was first put
forward by . orovi Ljubinkovi, Odnosi Slovena centralnih oblasti
Balkana i Vizantije od VII do II, Materijali 9 (1972) 81, 89.
, , 391, 395.
It has been assumed that the Slavic fibulae discovered in skeleton graves in
Velesnica, Prahovo and Korbovo could be attributed to Christians: ,
. 25. There are lead crosses from the late seventh century among
the accidental findings from Kostolac, from the vicinity of Jagodina, and
from Vina; these indicate that the population, dwelling along the Danube
valley, was most likely Christianized.
, , 388-392.
, 118. He made this conclusion on the basis of the
results of the excavations conducted in the region of Novi Pazar.
218
Dejan Buli
After a quarrel with his brothers, the reign of the Serbian archont
Mutimir (851-891) went on rather peacefully. Only after his death, at the
end of the ninth century, the struggle for power began, followed by
undisturbed reign of Peter that ended when Simeon captured him in 917,
in the aftermath of the battle of Anchialus.842 Simeon installed Paul
(Pavle), who ruled for the following six years (918924), and was followed
by Zacharius rise to power and the Bulgarian raid into Serbia, which
happened in 926, most likely.843 This was a year of great destruction,844 and
of an apparent gap in the reign of Serbian archonts. At this point,
fortifications were temporarily abandoned, until aslav took over the
power in Serbia (933-943).845 Belo, one of aslavs successors, was forced to
fight another war with the Syrmians and the Hungarians and won the
battle of Belina (Bellina).846 Although many toponyms bear that name,
there is a river crossing across the Sava in the present-day Mava that even
today has that name. Gradutina (hydroelectric power station Gradutina),
near Beljina (90) and Kupinovo were, in all likelihood, fortresses built on
the crossing point across the Sava, indicated by the toponym of the nearby
village Skela (meaning ferry). According to the Kanics sketch, the
bridge, whose remains are still visible, was on the road that passed through
a fort. Gradutina and Kupinovo are on the road connecting Bassianae
(Bassianae Donji Petrovci) (159) and Cusum (Cusum - Petrovaradin).847
Archaeological excavations confirmed the existence of layers dating until
the end of the twelfth century.848 But the dilemma remains whether the
lands around the Sava in Mava (Mavansko Posavlje) were part of Serbia
just like the lands around the Sava in Bosnia were (Bosansko Posavlje);
having in mind the account of the Priest of Duklja on the common struggle
of the Hungarians and the Syrmians.849
842
843
844
845
846
847
848
849
Ibid, 413.
Ibid, 421
DAI I, 32.119-126; II, 56.
On the years of the reign and the territory of the state, cf. . ,
, 55.
Moin, Ljetopis, 72.
I. Popovi, Notes topographiques sur la rgion limitrophe entre la Pannonie
Seconde et la Msie Premire, Roman Limes on the iddle and Lower
Danube, 1996, 137-142.
I, ,
1953, 16 (. - . ).
, , 432.
219
In the chapter XXXII of his treatise De administrando imperio,

Porphyrogenitos mentions the following inhabited towns in Serbia:
Destinik, ernavusk, Meureje, Dresneik, Lesnik and Salines, and Kotor
and Desnik as inhabited towns of Bosnia.850 The differing locations of the
towns Porphyrogenitos mentions as Serbian indicate what an inaccurate
and fanciful thing it is. According to S. Novakovi, the first Serbian towns
were situated in the eastern and south-eastern parts of the land, towards
the Ibar river and Bulgaria, the source of danger for the country.851 The
events within the ruling dynasty and the flights of individuals in
neighbouring lands, primarily Croatia, indicate that the seats of the
archonts were located somewhere in the west. The location of the meeting
between Petar Gojnikovi and the strategos of Dra, that took place in the
lands of Arentani (Narentines), points to the same conclusion. It can be
assumed that the Serbs protected their eastern borders from the
Bulgarians, while the rulers continued to govern the land from the west of
the country.852 Although the locations of the mentioned towns were
proposed back in the nineteenthth century, most of these sites were not
archaeologically surveyed. The only assumptions are that Destinik could
be at Vrsenice (97),853 where there are layers of this period; and Lesnik,
which S. Novakovi considered to be at Ljenica by Vidojevica (46),854
where findings of tenth- to twelfth-century pottery were discovered.855
But by the beginning of the eleventh century, not one of these
towns was mentioned in the bulls issued to the Archdiocese of Ohrid by
Basil II, suggesting that these towns were either fortifications on rather
inaccessible terrain, unsuitable for permanent settling, or that they began
to lose their importance for reasons unknown to us. If these were part of a
chain of fortifications along the Serb-Bulgar border, they became obsolete
with Samuils conquests and fell into disuse.856
850
851
852
853
854
855
856
DAI I, 32.149-151; II, 58.

. , II , 48 (1880) 140-143.
, 121.
Popovi - Biki, Vrsenice, 134. The toponym Crni vrh is situated in the
nearest surroundings of this site.
. , X XII ,
(. . ), 2003, 203.
. Jankovi, Rekognosciranje srednjovekovnih nalazita u zapadnoj Srbiji i
na Peteru, AP 20 (1978) 186. For further information on the preserved
remains of the town, see:
I, , 1953, 45-46 (. - . ).
, 125.
220
Dejan Buli
Very soon after the Hungarians arrived to Pannonia, Great

Moravia was destroyed; masses of refugees fled to the neighbouring lands
from the Hungarian marauders. Typical Moravian findings confirm this
assumption: an axe, discovered close to Vrac857, a vessel from Poarevac,
jewellery from Ram (163) and from Karaburma, most probably brought by
the refugees,858 and the well-known finding from Trilj.859
The afore-mentioned Hungarian incursion instigated the
foundation of new towns in Bulgaria (unfortified Ram with accidental
findings, Veliko Gradite (164)?, Veliki Gradac (170), Tekija (135)?,
Trajanov Most?-Kladovo (126)?, Prahovo (129); several smaller fortresses
had been restored, like the one situated on a hill overlooking the Poreka
river (168), and a new tower was erected on the Early Byzantine fortress
near the Dunav resort (121).860
The Hungarians took advantage of the succession on the Bulgarian
throne after the death of Simeon (893-927) and penetrated into Bulgaria
all the way to Macedonia and to the Black Sea. As a result, several
settlements on the Danube were abandoned, such as Dunav-Slatina
(121), the settlement in Mihajlovac (123), the fortification above the
Poreka river, but also settlements that were located outside the forts, like
Fetislam. Most likely, the jewellery hoard from Boljetina (166) was cached
at this point.861 After the Hungarian arrival, many Slavic refugees were
welcomed in the neighbouring lands. Findings of the Belobrdo culture
have been discovered deep in the Balkan hinterland, even in Kosovo and
the coast, where refugees brought them.862 Around this time, aslav fought
the Hungarians in the west.
Byzantium made good use of the military weakening of Bulgaria
and managed to retake positions on the Danube, down the stream from
erdap, with the help of the Russians and the prince Svyatoslav (946-972).
The abandonment of an unfortified site near Grabovica, and the ending of
857
858
859
860
861
862
. , , 1977, 16-17.
. . , , 36.
P. Koroec, Kronoloka i kulturna ocjena triljskog nalaza, SP 21 (1991)
1995/96, 87-96
. . , , 42.
. . , , 37.
. ,
,
, 1988, 25.
221
use at a necropolis by Trajans bridge near Kladovo, could be linked with

the Russian military campaign.863
In the early eleventh century the restoration of the Byzantine
power came about with the demise of Samuils empire. After the new
administrative order and ecclesiastical organization had been established,
some towns experienced revival and gained new functions. Military crews
retook strategically forefront positions, such as frontiers, roads and
administrative centres. Along with the new ecclesiastical organization, the
role of old episcopal towns became more important within the frame of
the Archdiocese of Ohrid that the emperor brought under his authority.
The Byzantine tradition took the ecclesiastical organization of the
Prefecture of Illyricum as a foundation when, through the bulls of Basil II
(976 - 1025), the authority of the Archbishop of Iustiniana Prima was
transferred onto the bishop, that is to say, the Archbishop of Ohrid.864 In
the already-mentioned bulls issued by Emperor Manuel to the
Archdiocese of Ohrid, transcriptions were included of the bulls of 1019,
May 1020 and from 1020-1025, issued by Basil II to the same church. The
bull of 1019 lists 17 dioceses in total, six of which were in Serbia: Ni,
Branievo, Beograd, Sirmium, Prizren and Lipljan. Next to every episcopal
see were listed towns in its demesne, with the number of clerics and
parishioners written down. In the second bull, another 14 dioceses were
attached to the Archdiocese of Ohrid, raising the total sum to 31, of which
only Ras was in the territory of the present-day Serbia.865
Here is the list of dioceses in the territory of the present-day Serbia
(including towns in their demesne):
Ni, with the following towns: Mokro (Bela Palanka), Kabl (lying
on the road Prokuplje-Ni), Toplica (Kurumlija), Sfeligovo (Svrljig);
Branievo, with the following towns: Moravisk (Morava),
Sfeneroman (Smederevo? or its surroundings), Grocka or Grua, Divisisk
(Leve or Temni), Stala, Brodarisk (uprija);
Prizren: Hosno (or the region Hvosno), Leskovac (at the location
where the Knina joins the Drim), Vret (Brut or Vrmnica, both lying southwest of Prizren);
863
864
865
. . , , 38.
, , 37.
For more details on this issue, including the map of the dioceses and the
towns within their jurisdiction, see: , ,
172-177.
222
Dejan Buli
Belgrade, with the following towns: Gradein (Gradac near

Valjevo?), Omcon (Uice?), Glavenica (?) and Bela Crkva (?);
In the demesnes of Dmitrovica, Lipljan and Ras, no towns.866
Without venturing into unreliable hypothetical locations, only the
towns in the list are mentioned, i.e. the towns with foundations in the
Antiquity, and of familiar positions: Ni (175), Morava (162), Branievo
(8), uprija (120), Prizren (225), Belgrade (161), Ras.
The throne of the Bishops of Ras is normally taken to have been in
the church of St. Apostles Peter and Paul (St. Peters church by Novi
Pazar), while the town of Ras was probably situated at Gradina-Postenje
(116)867, which corresponds to the discovered material, although locating it
at Gradina by Trgovite (111) is not without grounds.868
During Samuils reign, the Serbs were certainly not allowed to use
or maintain any fortified strongholds, but when Basil threatened him,
Samuil must have made some fortifications ready for defence. The reign of
Basil II had no use of the high fortifications, except those that served the
needs of Byzantine authorities. After the death of Basil II, the economic
situation became worse. The increasing taxes and the introduction of taxes
payable in money, led to an uprising in 1040, which spread across all Slavic
lands as far as Thessaly (Margum, Belgrade, Ni, Skoplje). The uprising was
quelled shortly after, in 1041.
Fortifications could also have been used during the uprisings in the
eleventh century in 1040, 1072, etc. The fortification of Belgrade was
mentioned in the historical sources recounting the Byzantine-Hungarian
war of 1071.869 It would be naive to think that the uprising which broke
out a year later had no correlation to these events. And except for
demonstrating the desire to get rid of the Byzantine rule, this uprising
shows that some fortifications were most likely used, at least those
overlooking the roads and suitable for organizing surprise attacks on
smaller Byzantine military units. After the uprising of 1071, the Byzantine
866
867
868
869
For further information on locating the positions of these towns, see: .

, XI .
II 1019. 1020. ,
(. . ), 2003, 61-102; I,
1983, 178 (. ); . , ,
160; , , 176.
. , -. , 36 (1997) 203-217.
. Popovi, Tvrava Ras, Beograd 1999.
. , I, 1951,
11-12.
223
authorities seem to have resettled the population from the fortified

Danubian towns since several excavated fortifications were abandoned at
the same time, around 1072. These were: Veliki Gradac, Tekija, Trajans
bridge, Prahovo and Gamzigrad. In Prahovo, additionally corroboration
comes from the absence of coins minted after this year. Traces of life
reappear only at Trajans bridge (Kladovo), but with a new population that
brought new pottery with them.870
Conclusions
In the Late Antiquity, fortifications sprang up all over the Roman

Empire. Beside the restoration of the fortifications on the Danubian limes,
which was the pinnacle of military architecture in the Antiquity, the focus
was on organizing defence in depth, to prevent or at least buffer barbarian
incursions into the interior of the Balkan Peninsula and soften the blow on
the great urban centres of the Mediterranean. The hill forts are not
particularly distinguishing for their fortifications; rather, their key
advantage, in a military and strategic sense, was the inaccessible and
naturally defensive terrain that did not require the construction of strong
and complex fortifications. And while the forts on the limes, the forts along
the traffic ways, and those in the mining districts were part of a singular
defensive system, a large number of the fortifications in the interior were
solely dedicated to securing local or regional defence.
One of the objectives of this work was to compile the lists and the
maps of the Late Antiquity/Early Byzantine fortifications that would
provide a sound basis for further research. But it was also our aim to reflect
on the wider historical context in which these fortifications came to be;
and to do so to the degree the current state of research of these sites allows
us, not to mention the specifics of particular fortifications, construction
technique, movable and immovable findings, the functions of fortifications,
their mutual relations and their role in the defensive system of the Empire.
Procopius gives us a total figure of 654 fortifications in the territory
of the Balkan Peninsula, but the figure of those known to us surpasses this
number by far.871 Approximately 500 fortifications were located in the
870
871
. . , , 40.
. , . , 2010, 226.
224
Dejan Buli
territory of Macedonia, 259 in Serbia, more than 300 in Bosnia and

Herzegovina, almost 100 in Croatia, and at least 17 in Montenegro,
without thorough survey or research, and without counting the multitude
of fortifications in Bulgaria, Greece, and Albania. That is to say, just in this
work, over a thousand fortifications have been encompassed, in a way.
With all the deficiencies of such a classification, the number of
classified fortifications still does not correspond to the actual figure,
because of the poor surveys in some areas. But it is getting closer to the
actual number. The empty zones are not there because they were
uninhabited in the Late Antiquity, but because of the insufficient research
that has been carried out. Also, in some areas, several gradinas were, with
inertia, designated as prehistoric. But in time, with the progress of research,
it should be expected that the empty zones will be filled out with new sites.
Observing one long period as economic crisis starting in the third
century, might not be the happiest of solutions, since such a long period
could be defined as a state, rather than as a crisis. Perpetual barbarian
attacks led to the gradual evacuation of the northern parts of Illyricum, i.e.
the most threatened regions. A point has already been raised, of two
directions of migrations vertical and horizontal. As the state lasted, the
transfer to the locations difficult to access was carried out completely and the
flatland expanses of Illyricum were abandoned, given the fact that the
settlements of the sixth century are unknown to have existed in the flatlands,
except for the fortifications. In contrast to this, an entire web of singlelayered fortifications sprang up in the high terrain. The horizontal migrations
led to migrations towards south the population withdrew to the coast and
to the islands, where the water provided the only traffic way possible. We
can assume that the wealthier kept on withdrawing deeper southwards, into
safety, while the poor remained most exposed to the attacks.
The issue of depopulation was certainly less of a problem on the
coast, because of the constant influx of refugees from the north. They
probably provided a cheap workforce, new craft skills and entrepreneurial
spirit. Some of them, of the wealthier kind, must have brought money and
provided a financial injection for the littoral belt. We are not about to say
that things were blossoming at the time, but generalisation of the urban
environments dying off is not entirely accurate either, as we have seen
from the process of creation of new settlement both on the coast and in the
interior. Maybe it would not be wrong to observe the process of castrization
225
as urbanisation, to a certain degree, because, aside from the populations

from the plains that came to the highlands, the pastoral population that
probably lived in scattered villages was now brought into a relatively cramped
space, which demanded a certain organisation and fulfilment with
different contents and buildings, most often churches and workshops.
The final objective of this work was to register the early medieval
and medieval strata at the existing fortifications, and to determine if there
was continuity and/or discontinuity in the medieval and Early Byzantine
period. Unlike the coastal towns of Dalmatia that continuously lasted,
there is almost not a single site in the interior that was settled immediately
after the arrival, except for short-term use. This situation shows
discontinuity of fortifications and it demonstrates that the re-use came
about as a consequence of new historic conditions.
In Bosnia, there are 41 Late Antique/Early Byzantine sites with
medieval traces. Considering there are in total 319 Late Antique/Early
Byzantine sites, this equals 12.81%. In Macedonia, the percentage is
slightly higher, where a medieval town or fortification arose on 16.4
percent of the fortification sites from the Antiquity. It can be deduced
from this that in medieval Macedonia newly-erected medieval stone
fortifications were quite scarce and were more of an exception than a rule.
We do not have a good insight in Croatia, but out of 89 fortifications, 16
were re-used, equalling 17.89%. In Montenegro, out of 17 sites, 11 have
later phases of occupation, or 64.7%. In Serbia, 259 sites have been
registered, out of which 84 re-used, amounting to 32.43%.
We hold this percentage to be much higher in reality, having in
mind the already mentioned flaws and scarcity of information that make
the isolation of the medieval layers impossible. The most accessible and
accurate information pertains to Serbia, considering the poor knowledge
of most sites and them being merely registered, in most of the cases. That
is why these data would be most faithful to the actual situation in the field.
Still, we expect that, with increased insight and research of the
fortifications, the percentage of those that had been reutilized will rise to
35%, and possibly even to 40%.
The architecture of the fortifications from the period of their
medieval re-use did not differ from the architecture in the Antiquity. In
the territory of Illyricum, the Slavs encountered a multitude of fortifications
that had much of their ramparts and towers in sound condition. Minimal
226
Dejan Buli
reconstruction works on the wall battlements can no longer be seen

today, since those parts of walls are gone. The wooden superstructures,
partitions and dwellings made of light materials have either not been
preserved or, rarely, have been preserved in traces. This is one of the
reasons for inability to recognize the architectural elements that could be
attributed to the Slavs. The only evidence of a sometime use are frequently
movable findings, of which a good part is impossible to chronologically
determine, such as tools. On the other hand, pottery was never given due
attention, or reliable differentiation by epochs. This will call, in the
foreseeable future, for a revision and re-dating of some pottery fragments
from certain sites.
After the temporary use of particular fortifications during the
seventh century, the previously defended space went desolate and
uninhabited for more than two centuries. The absence of findings speaks
of these forts lying vacant. The high-altitude terrain did not appeal to the
Slavic tribes, which is why the traces of their presence should be looked
for in valleys and river basins, until the ninth- and tenth- century phase of
re-use, caused by wars and the need of their use. The only exception, for
now, is Gradina on Jelica. The encountered fortifications were partly
reconstructed with minimal interventions. On the other hand, thorough
reconnaissance of flatland positions never took place, at least not in a way
that would enable identification of flatland settlements.
Throughout the historical epochs and challenges, well-tried
strategical positions were re-used in the Middle Ages, too. But then as seats
of nobles and lords, seats of upas (districts), or as important frontier forts.
During the Ottoman rule, very few of these points were used again, since
garrisons stationed in the borderline areas, along most important traffic
ways, and near economic centres, sufficed. The Turks maintained only the
most important fortifications after the conquest, while the others were
dismantled, preventing their later use.
Archaeological findings speak of the relationship the Slavs built for
their new environment, but the use of these sites is not a proof of the
newly-arrived population adjusting to the previous settlements, nor is it a
proof of the continuity of life. Rather, it is about the analogous factors
leading to their subsequent re-use, which is the state of immediate war
danger. Chosen with foresight and situated on important points, they
justified the decision to build them with the strategic role they played and
227
the importance they held down through the centuries. Only a few points
on the coast and the islands remained continuously inhabited, where the
continuity was upheld by the autochthonous population, which in time
included into their ethnic group the fresh blood from the Slavic hinterland.
228
T. 1 Roman Fortifications in Bosnia (by I. remonik, with the addition by D. Buli)
229
T. 2 Early Byzantine Fortifications in Croatia
230
T. 3 Early Byzantine Fortifications in Montenegro
T. 4/1 Early Byzantine Fortifications in Macedonia (by I. Mikuli)
231
T. 4/2 Medieval Towns and Other Forts in Macedonia (by I. Mikuli, with the addition by D. Buli)
232
233
T. 5 Early Byzantine Fortifications in Serbia

Dejan Bulić, The Fortifications of The Late Antiquity and The Early Byzantine Period On The Later Territory of The South-Slavic Principalities, and Their Re-Occupation

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Dejan Bulić, The Fortifications of The Late Antiquity and The Early Byzantine Period On The Later Territory of The South-Slavic Principalities, and Their Re-Occupation

Hochgeladen von

Copyright:

Verfügbare Formate

THE WORLD OF THE SLAVS

Studies on the East, West and South Slavs:

This book has been published with the financial support of

THE INSTITUTE OF HISTORY

Tibor ivkovi Dejan Crnevi, Dejan Buli

THE WORLD OF THE SLAVS

The Urban Landcape of Early Medieval Slavic Principalities in

The Architecture of Cathedral Churches on the Eastern Adriatic

The Fortifications of the Late Antiquity and the Early Byzantine

Terrestrial Communications in the Late Antiquity and the Early

The Typology of Early Medieval Settlements in Bohemia, Poland

The Typology of Slavic Settlements in Central Europe in the

The Fortifications of the Late Antiquity and the Early

M. Garaanin, Odbrambeni sistemi u praistoriji i antici na tlu Jugoslavije,

Issues regarding precise delineations of the Late Antiquity provinces have

The Fortifications of the Late Antiquity and Early Byzantine Period

The Province of Dalmatia A Historical Overview

With the Hunnic invasion, the majority of Illyrian towns were

For further information on the history of towns in the mid-400s Illyricum,

policy of Justinian I, he strived to reshape the borders of archdioceses so as

Bosnia and Herzegovina:

Bosnia and Herzegovina occupies the central part of the Balkans.

The Fortifications of the Late Antiquity and Early Byzantine Period

Patsch, Zbirka, 159.

The Fortifications of the Late Antiquity and Early Byzantine Period

264. Gradina, Rajike, Glamo210

, , 42; Leksikon 3, 240; A. Benac, Utvrena ilirska

273. Mali Grad-Blagaj kod Mostara.220

. Basler, Arhitektura kasnoantikog doba u Bosni i Hercegovini, Sarajevo

The Fortifications of the Late Antiquity and Early Byzantine Period

287. Glavica, Mali Mounj, Vitez234

Koroec, Travnik, 257; Leksikon 2, 197.

300. Velika Gradina, Slivnica, Trebinje247

. Odavi, Praistorijska nalazita na prostoru Trebinja (gomile i gradine),

The Fortifications of the Late Antiquity and Early Byzantine Period

314. Vukove Njive, Gradac, Posuje261

Procopius makes no mention of fortification construction in Dalmatia.

The Fortifications of the Late Antiquity and Early Byzantine Period

remonik, Rimska utvrenja, 357.

comes as a consequence of geographic conditions i.e. the fact that these

The Fortifications of the Late Antiquity and Early Byzantine Period

6. Mogorjelo, apljina (252)283

15. Bobovac, Dragovii-Miljakovii, Vare (63)292

The Fortifications of the Late Antiquity and Early Byzantine Period

24. Glavica, Mali Mounj, Vitez (287)301

Koroec, Travnik, 257; Leksikon 2, 197.

35. Mali Grad-Blagaj near Mostar (273)312

The Fortifications of the Late Antiquity and Early Byzantine Period

medieval settlements; site 19 as a medieval building, while individual

Constantine Porphyrogenitus, De administrando imperio I (ed. Gy.

fortifications, fragments of early Slavic ceramics were discovered. These

The Fortifications of the Late Antiquity and Early Byzantine Period

Most of the present-day Croatia belonged to the province of

M. Sanader, Rimske legije i njihovi logori u hrvatskom dijelu panonskog

The Fortifications of the Late Antiquity and Early Byzantine Period

Andetrium and Bigeste.334 After the conquest of Dalmatia, the population

D. Peria, Je li delmatsko podruje presjekao rimski limes?, Archaeologia

fications allegedly existed (but were yet to be confirmed), some positions

Such assumptions are supported, in some cases, by the toponyms of these

The Fortifications of the Late Antiquity and Early Byzantine Period

10. Polaa in ljet349

M. Sui, Antiki grad na istonom Jadranu, Zagreb 1976, 239.